ADAPTATION OF SCAN ORDER FOR ENTROPY CODING

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
0Forward
Citations 
0
Petitions 
0
Assignments
First Claim
1. An apparatus for decoding coefficients of a transform block, comprising:
 a processor, the processor is configured to;
decode, from a compressed bitstream, first coefficients of at least one row of the transform block;
decode, from the compressed bitstream, second coefficients of at least one column of the transform block, the transform block consisting of the first coefficients, the second coefficients, and remaining coefficients;
select, using the first coefficients and the second coefficients, a scan order for decoding the remaining coefficients; and
decode the remaining coefficients of the transform block using the scan order.
0 Assignments
0 Petitions
Accused Products
Abstract
Apparatuses and methods are disclosed for adapting scan order of a transform block. An apparatus for decoding coefficients of a transform block includes a processor. The processor is configured to decode, from a compressed bitstream, first coefficients of at least one row of the transform block; decode, from the compressed bitstream, second coefficients of at least one column of the transform block, where the transform block consists of the first coefficients, the second coefficients, and remaining coefficients; select, using the first coefficients and the second coefficients, a scan order for decoding the remaining coefficients; and decode the remaining coefficients of the transform block using the scan order.
0 Citations
No References
No References
20 Claims
 1. An apparatus for decoding coefficients of a transform block, comprising:
a processor, the processor is configured to; decode, from a compressed bitstream, first coefficients of at least one row of the transform block; decode, from the compressed bitstream, second coefficients of at least one column of the transform block, the transform block consisting of the first coefficients, the second coefficients, and remaining coefficients; select, using the first coefficients and the second coefficients, a scan order for decoding the remaining coefficients; and decode the remaining coefficients of the transform block using the scan order.  View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9)
 10. A method for decoding coefficients of a transform block, comprising:
decoding, from a compressed bitstream, first coefficients of at least one row of the transform block; decoding, from the compressed bitstream, second coefficients of at least one column of the transform block, the transform block consisting of the first coefficients, the second coefficients, and remaining coefficients; selecting, using the first coefficients and the second coefficients, a scan order for decoding the remaining coefficients; and decoding the remaining coefficients of the transform block using the scan order.  View Dependent Claims (11, 12, 13, 14, 15, 16, 17, 18)
 19. An apparatus for encoding coefficients of a transform block, comprising:
a processor, the processor is configured to; encode, in a compressed bitstream, first coefficients of at least one row of the transform block; encode, in the compressed bitstream, second coefficients of at least one column of the transform block, the transform block consisting of the first coefficients, the second coefficients, and remaining coefficients; select, using the first coefficients and the second coefficients, a scan order for encoding the remaining coefficients; and encoding the remaining coefficients of the transform block using the scan order.  View Dependent Claims (20)
1 Specification
This application claims priority to and the benefit of U.S. [Provisional] application patent Ser. No. 15/608,008, filed May 30, 2017, the entire disclosure of which is hereby incorporated by reference.
Digital video streams may represent video using a sequence of frames or still images. Digital video can be used for various applications including, for example, video conferencing, high definition video entertainment, video advertisements, or sharing of usergenerated videos. A digital video stream can contain a large amount of data and consume a significant amount of computing or communication resources of a computing device for processing, transmission or storage of the video data. Various approaches have been proposed to reduce the amount of data in video streams, including compression and other encoding techniques.
Encoding based on spatial similarities may be performed by breaking a frame or image into blocks that are predicted based on other blocks within the same frame or image. Differences (i.e., residual errors) between blocks and prediction blocks are compressed and encoded in a bitstream. A decoder uses the differences and reference frames to reconstruct the frames or images.
This application relates to encoding and decoding a current block of video data. Disclosed herein are aspects of systems, methods, and apparatuses for encoding and decoding the current block using adaptation of scan order for entropy coding.
One aspect of the disclosed implementations is an apparatus for decoding coefficients of a transform block. The apparatus includes a processor. The processor is configured to decode, from a compressed bitstream, first coefficients of at least one row of the transform block; decode, from the compressed bitstream, second coefficients of at least one column of the transform block, where the transform block consists of the first coefficients, the second coefficients, and remaining coefficients; select, using the first coefficients and the second coefficients, a scan order for decoding the remaining coefficients; and decode the remaining coefficients of the transform block using the scan order.
Another aspect is method for decoding coefficients of a transform block. The method includes decoding, from a compressed bitstream, first coefficients of at least one row of the transform block; decoding, from the compressed bitstream, second coefficients of at least one column of the transform block, where the transform block consists of the first coefficients, the second coefficients, and remaining coefficients; selecting, using the first coefficients and the second coefficients, a scan order for decoding the remaining coefficients; and decoding the remaining coefficients of the transform block using the scan order.
Another aspect is an apparatus for encoding coefficients of a transform block. The apparatus includes a processor. The processor is configured to encode, in a compressed bitstream, first coefficients of at least one row of the transform block; encode, in the compressed bitstream, second coefficients of at least one column of the transform block, where the transform block consists of the first coefficients, the second coefficients, and remaining coefficients; select, using the first coefficients and the second coefficients, a scan order for encoding the remaining coefficients; and encoding the remaining coefficients of the transform block using the scan order.
These and other aspects of the present disclosure are disclosed in the following detailed description of the embodiments, the appended claims and the accompanying figures.
The description herein makes reference to the accompanying drawings wherein like reference numerals refer to like parts throughout the several views.
Compression schemes related to coding video streams may include breaking images into blocks and generating a digital video output bitstream (i.e., an encoded bitstream) using one or more techniques to reduce the data rate of the output bitstream. A received bitstream can be decoded to reconstruct the blocks and the source images from the compressed information. Techniques used to encode and decode video streams can include entropy coding of components of a video signal (e.g., a prediction residual for a block from a frame of video).
A hybrid video coding framework may make use of prediction, transform, quantization, and/or entropy coding to compress a block of video data. Depending upon a pattern occurring in a block of quantized transform coefficients after prediction, transform, and quantization, compression performance of entropy coding may be improved by adaptively selecting a scan order instead of using a fixed scan order. For example, if a significant number of zero valued coefficients occur at the end of the scan order for the block, then the ending run of zeros can be very efficiently coded (e.g., using an entropy encoder or an entropy decoder) explicitly or implicitly (e.g., by an end of block indication).
A block may be partitioned for entropy coding into a plurality of groups of elements (e.g., quantized transform coefficients of a residual block), and information in the first group of elements that are coded using a default scan order, which is available at both the encoder and compatible decoder, may be used to determine a scan order for a next group. For example, based on a distribution of nonzero valued elements in prior coded groups of elements, a scan order for the last group(s) of elements of the block may be selected to increase the probability of zero valued elements occurring at the end of the resulting scan order for the block. In some implementations, scan order adaptation is based only on information available in the current block of elements (e.g., quantized transform coefficients of a residual block). Such a scheme may offer the advantage of avoiding dependency of the entropy coding of the block on other parameters, such as an intraprediction mode for the block or coefficient statistics for other blocks of the video that may be used by a constrained adaptive scan order scheme.
By adaptively determining the ending scan order for a block based on information in a partitioned group of elements of the block that start the scan order, a number of advantages may be achieved over other coding schemes. For example, the advantages may include (1) increasing the coding efficiency of the an entropy encoder/decoder by using a scan order that is adapted to the distribution of nonzero coefficients in the current block; (2) allowing the processing of blocks of a frame in different orders by avoiding a parsing dependency on information outside of the current block; and/or (3) enabling the improvement of intraprediction mode coding efficiency by enabling intraprediction modes to be decoded after the elements of a residual block and leveraging correlation between the elements of the block and the intraprediction mode used to generate the block.
The scan order for a block based on a residual block resulting from interprediction may be adapted based on information in a partitioned group of elements of the block that start the scan order. For example, interprediction may use a motion vector that represents the spatial displacement of a previously coded block relative to the current block. The motion vector can be identified using a method of motion estimation, such as a motion search. In the motion search, a portion of a reference frame can be translated to a succession of locations to form a prediction block that can be subtracted from a portion of a current frame to form a series of residuals. The horizontal and/or vertical translations corresponding to the location having, e.g., the smallest, residual can be selected as the motion vector. The motion vector can be encoded in the encoded bitstream along with an indication of the reference frame. A resulting interprediction residual may exhibit spatial patterns that can be recognized and exploited to improve entropy coding efficiency for the residual by adapting the scan order for a block based on the residual.
Encoding a video stream, or a portion thereof, such as a frame or a block, can include using spatial similarities in the video stream to improve coding efficiency. For example, a current block of a video stream may be encoded based on identifying a difference (residual) between the previously coded pixel values, or between a combination of previously coded pixel values, and those in the current block.
Encoding using spatial similarities can be known as intra prediction. Intra prediction attempts to predict the pixel values of a current block of a frame of a video stream using pixels peripheral to the current block; that is, using pixels that are in the same frame as the current block but that are outside of the current block. Intra prediction can be performed along a direction of prediction, referred to herein as prediction angle, where directions can correspond to respective intraprediction modes. An intraprediction mode uses pixels peripheral to the current block being predicted. Pixels peripheral to the current block are pixels outside the current block.
Many different intraprediction modes can be supported. Some intraprediction modes may be referred to as directional intraprediction modes, which may have corresponding prediction angles. Directional intraprediction modes can include, for example, horizontal intraprediction modes, vertical intraprediction modes, and diagonal intraprediction modes. A horizontal intraprediction mode corresponds to an intraprediction having a prediction angle that is generally horizontal. A vertical intraprediction mode corresponds to an intraprediction having a prediction angle that is generally vertical. A diagonal intraprediction mode corresponds to a directional intraprediction having a prediction angle that is neither generally horizontal nor generally vertical. The prediction angle of a directional intraprediction mode can be a prediction angle between 0 and 360 degrees. In some implementations, the prediction angle can be a prediction angle between 0 and 270 degrees. Available prediction angles can also be a subset of all possible prediction angles. For example, a codec can have available prediction modes corresponding to 5060 discrete prediction angles of the 0 to 360 prediction angles. Non directional intraprediction modes can also be supported, including, for example, a planar intraprediction mode and a DC intraprediction mode.
Directional intraprediction modes can be used to propagate pixel values from previously coded blocks along an angular line to predict a block. For example, pixel values being propagated can include peripheral pixels above and/or to the left of the block in the same frame (e.g., when raster scan order is used in encoding).
As further described below, a residual block of residual errors can result from the intra prediction of a block of a video stream. The residual block can be in the pixel domain. The residual block can be transformed into the frequency domain resulting in a transform block of transform coefficients. The transform coefficients can be quantized resulting into a quantized transform block of quantized transform coefficients (also referred to herein as quantized coefficients). A quantized transform block of quantized transform coefficients is also referred to herein as a block of quantized transform coefficients. The quantized coefficients can be entropy encoded and added to an encoded bitstream. The intraprediction mode can also be entropy encoded and added to the encoded bitstream. A decoder can receive the encoded bitstream, entropy decode the quantized transform coefficients and/or the intraprediction mode to reconstruct the original video frame.
The number of bits used to encode the intraprediction mode can depend on the number of available intraprediction modes. In a video coding system where N intraprediction modes are available, log_{2 }N bits may be required to indicate the prediction mode. For example, in a system where nine (9) or ten (10) intraprediction modes are available for a 4×4 luminance block, four (4) bits may be required to indicate an intraprediction mode. In a system where 35 intraprediction modes are available for a 4×4 luminance block, six (6) bits may be required to indicate an intraprediction mode. Coding systems that use more intraprediction modes require more bits to encode an intraprediction mode.
As the number of available intraprediction modes increases, the cost (in bits) of encoding an intraprediction mode increases. In some situations, the number of bits required to encode an intraprediction mode can be greater than the number of bits required to encode the pixel values of the current block being encoded. For example, to encode a 4×4 block (which contains 16 pixels) in a coding system that uses more than 16 intraprediction modes, the number of distinct prediction modes is higher than the number of pixels to be predicted. This results in a problem of over saturation.
Implementations of this disclosure can improve video compression using improved coding of intraprediction modes. Implementations according to this disclosure can use information in residual blocks to improve the compression performance of intraprediction modes. For example, information in the quantized transform block can be used by an entropy encoder to encode the intraprediction mode thereby improving compression performance. By using information in the quantized transform block, the bit rates resulting from encoding an intraprediction mode can be reduced. Encoders and decoders according to implementations of this disclosure can improve the entropy coding without increasing the complexity of the encoding and decoding.
Details are described herein after first describing an environment in which the improved coding of intraprediction modes disclosed herein may be implemented.
A network 104 can connect the transmitting station 102 and a receiving station 106 for encoding and decoding of the video stream. Specifically, the video stream can be encoded in the transmitting station 102 and the encoded video stream can be decoded in the receiving station 106. The network 104 can be, for example, the Internet. The network 104 can also be a local area network (LAN), wide area network (WAN), virtual private network (VPN), cellular telephone network or any other means of transferring the video stream from the transmitting station 102 to, in this example, the receiving station 106.
The receiving station 106, in one example, can be a computer having an internal configuration of hardware such as that described in
Other implementations of the video encoding and decoding system 100 are possible. For example, an implementation can omit the network 104. In another implementation, a video stream can be encoded and then stored for transmission at a later time to the receiving station 106 or any other device having memory. In one implementation, the receiving station 106 receives (e.g., via the network 104, a computer bus, and/or some communication pathway) the encoded video stream and stores the video stream for later decoding. In an example implementation, a realtime transport protocol (RTP) is used for transmission of the encoded video over the network 104. In another implementation, a transport protocol other than RTP may be used, e.g., a HyperText Transfer Protocol (HTTP)based video streaming protocol.
When used in a video conferencing system, for example, the transmitting station 102 and/or the receiving station 106 may include the ability to both encode and decode a video stream as described below. For example, the receiving station 106 could be a video conference participant who receives an encoded video bitstream from a video conference server (e.g., the transmitting station 102) to decode and view and further encodes and transmits its own video bitstream to the video conference server for decoding and viewing by other participants.
A CPU 202 in the computing device 200 can be a central processing unit. Alternatively, the CPU 202 can be any other type of device, or multiple devices, capable of manipulating or processing information nowexisting or hereafter developed. Although the disclosed implementations can be practiced with a single processor as shown, e.g., the CPU 202, advantages in speed and efficiency can be achieved using more than one processor.
A memory 204 in the computing device 200 can be a readonly memory (ROM) device or a random access memory (RAM) device in an implementation. Any other suitable type of storage device can be used as the memory 204. The memory 204 can include code and data 206 that is accessed by the CPU 202 using a bus 212. The memory 204 can further include an operating system 208 and application programs 210, the application programs 210 including at least one program that permits the CPU 202 to perform the methods described here. For example, the application programs 210 can include applications 1 through N, which further include a video coding application that performs the methods described here. The computing device 200 can also include a secondary storage 214, which can, for example, be a memory card used with a computing device 200 that is mobile. Because the video communication sessions may contain a significant amount of information, they can be stored in whole or in part in the secondary storage 214 and loaded into the memory 204 as needed for processing.
The computing device 200 can also include one or more output devices, such as a display 218. The display 218 may be, in one example, a touch sensitive display that combines a display with a touch sensitive element that is operable to sense touch inputs. The display 218 can be coupled to the CPU 202 via the bus 212. Other output devices that permit a user to program or otherwise use the computing device 200 can be provided in addition to or as an alternative to the display 218. When the output device is or includes a display, the display can be implemented in various ways, including by a liquid crystal display (LCD), a cathoderay tube (CRT) display or light emitting diode (LED) display, such as an organic LED (OLED) display.
The computing device 200 can also include or be in communication with an imagesensing device 220, for example a camera, or any other imagesensing device 220 now existing or hereafter developed that can sense an image such as the image of a user operating the computing device 200. The imagesensing device 220 can be positioned such that it is directed toward the user operating the computing device 200. In an example, the position and optical axis of the imagesensing device 220 can be configured such that the field of vision includes an area that is directly adjacent to the display 218 and from which the display 218 is visible.
The computing device 200 can also include or be in communication with a soundsensing device 222, for example a microphone, or any other soundsensing device now existing or hereafter developed that can sense sounds near the computing device 200. The soundsensing device 222 can be positioned such that it is directed toward the user operating the computing device 200 and can be configured to receive sounds, for example, speech or other utterances, made by the user while the user operates the computing device 200.
Although
Whether or not the frame 306 is divided into the segments 308, the frame 306 may be further subdivided into blocks 310, which can contain data corresponding to, for example, 16×16 pixels in the frame 306. The blocks 310 can also be arranged to include data from one or more segments 308 of pixel data. The blocks 310 can also be of any other suitable size such as 4×4 pixels, 8×8 pixels, 16×8 pixels, 8×16 pixels, 16×16 pixels, 4×32 pixels, 8×32 pixels, 16×32 pixels, 32×4 pixels, 32×8 pixels, 32×16 pixels, 32×32 pixels, 64×64 pixels, or in general N×M pixels, where N, M may be an integer power of 2 like 2, 4, 8, 16, 32, 64, 128, 256, or larger.
When the video stream 300 is presented for encoding, the frame 306 can be processed in units of blocks. At the intra/inter prediction stage 402, a block can be encoded using intraframe prediction (also called intraprediction) or interframe prediction (also called interprediction), or a combination of both. In any case, a prediction block can be formed. In the case of intraprediction, all or a part of a prediction block may be formed from samples in the current frame that have been previously encoded and reconstructed. In the case of interprediction, all or part of a prediction block may be formed from samples in one or more previously constructed reference frames determined using motion vectors.
Next, still referring to
The quantization stage 406 converts the transform coefficients into discrete quantum values, which are referred to as quantized transform coefficients, using a quantizer value or a quantization level. For example, the transform coefficients may be divided by the quantizer value and truncated. The quantized transform coefficients are then entropy encoded by the entropy encoding stage 408. Entropy coding may be performed using any number of techniques, including token and binary trees. The entropyencoded coefficients, together with other information used to decode the block, which may include for example the type of prediction used, transform type, motion vectors and quantizer value, are then output to the compressed bitstream 420. The information to decode the block may be entropy coded into block, frame, slice and/or section headers within the compressed bitstream 420. The compressed bitstream 420 can also be referred to as an encoded video stream or encoded video bitstream, and the terms will be used interchangeably herein.
The reconstruction path in
Other variations of the encoder 400 can be used to encode the compressed bitstream 420. For example, a nontransform based encoder 400 can quantize the residual signal directly without the transform stage 404 for certain blocks or frames. In another implementation, an encoder 400 can have the quantization stage 406 and the dequantization stage 410 combined into a single stage.
When the compressed bitstream 420 is presented for decoding, the data elements within the compressed bitstream 420 can be decoded by the entropy decoding stage 502 to produce a set of quantized transform coefficients. The dequantization stage 504 dequantizes the quantized transform coefficients (e.g., by multiplying the quantized transform coefficients by the quantizer value), and the inverse transform stage 506 inverse transforms the dequantized transform coefficients using the selected transform type to produce a derivative residual that can be identical to that created by the inverse transform stage 412 in the encoder 400. Using header information decoded from the compressed bitstream 420, the decoder 500 can use the intra/interprediction stage 508 to create the same prediction block as was created in the encoder 400, e.g., at the intra/inter prediction stage 402. At the reconstruction stage 510, the prediction block can be added to the derivative residual to create a reconstructed block. The loop filtering stage 512 can be applied to the reconstructed block to reduce blocking artifacts. Other filtering can be applied to the reconstructed block. In this example, the deblocking filtering is applied by the postprocessing stage 514 to the reconstructed block to reduce blocking distortion, and the result is output as an output video stream 516. The output video stream 516 can also be referred to as a decoded video stream, and the terms will be used interchangeably herein.
Other variations of the decoder 500 can be used to decode the compressed bitstream 420. For example, the decoder 500 can produce the output video stream 516 without the postprocessing stage 514. In some implementations of the decoder 500, the postprocessing stage 514 (e.g., including deblocking filtering) is applied before the loop filtering stage 512. Additionally, or alternatively, the encoder 400 includes a deblocking filtering stage in addition to the loop filtering stage 416.
The quantized transform block 604 can be a block of size similar to the size of the current block 620. The quantized transform block 604 includes nonzero coefficients (e.g., a coefficient 608) and zero coefficients (e.g., a coefficient 610). As described above, the quantized transform block 604 contains quantized transform coefficients for the residual block corresponding to the current block 620. Also as described above, the quantized transform coefficients are entropy coded by an entropy coding phase, such as the entropy coding stage 408 of
Entropy coding a quantized transform coefficient can involve the selection of a context model (also referred to as probability context model, probability model, and context) which provides estimates of conditional probabilities for coding the binary symbols of a binarized transform coefficient. When entropy coding a quantized transform coefficient, additional information may be used as the context for selecting a context model. For example, the magnitudes of the previously coded transform coefficients can be used, at least partially, for determining a probability model.
To encode a transform block, a video coding system may traverse the transform block in a scan order and encode (e.g., entropy encode) the quantized transform coefficients as the quantized transform coefficients are respectively traversed (i.e., visited). In a zigzag scan order, such as the scan order 602, the top left corner of the transform block (also known as the DC coefficient) is first traversed and encoded, the next coefficient in the scan order (i.e., the transform coefficient corresponding to the location labeled “1”) is traversed and encoded, and so on. In the zigzag scan order (i.e., scan order 602), some quantized transform coefficients above and to the left of a current quantized transform coefficient (e.g., a tobeencoded transform coefficient) are traversed first. Other scan orders are possible. A onedimensional structure (e.g., an array) of quantized transform coefficients can result from the traversal of the twodimensional quantized transform block using the scan order.
In some examples, encoding the quantized transform block 604 can include determining the nonzero map 606, which indicates which quantized transform coefficients of the quantized transform block 604 are zero and which are nonzero. A nonzero coefficient and a zero coefficient can be indicated with values one (1) and zero (0), respectively, in the nonzero map. For example, the nonzero map 606 includes a nonzero 607 at Cartesian location (0, 0) corresponding to the coefficient 608 and a zero 608 at Cartesian location (2, 0) corresponding to the coefficient 610. In some cases, the nonzero map may be referred to as the significance map.
In some examples, encoding the quantized transform block 604 can include generating the endofblock map 622. The endofblock map indicates whether a nonzero quantized transform coefficient of the quantized transform block 604 is the last nonzero coefficient with respect to a given scan order. If a nonzero coefficient is not the last nonzero coefficient in the transform block, then it can be indicated with the binary value 0 (zero) in the endofblock map. If, on the other hand, a nonzero coefficient is the last nonzero coefficient in the transform block, then it can be indicated with the binary value 1 (one) in the endofblock map. For example, as the quantized transform coefficient corresponding to the scan location 11 (i.e., the last nonzero quantized transform coefficient 628) is the last nonzero coefficient of the quantized transform block 604, it is indicated with the endofblock value 624 of 1 (one); all other nonzero transform coefficients are indicated with a zero.
In some examples, encoding the quantized transform block 604 can include generating the sign map 626. The sign map 626 indicates which nonzero quantized transform coefficients of the quantized transform block 604 have positive values and which quantized transform coefficients have negative values. Transform coefficients that are zero need not be indicated in the sign map. The sign map 626 illustrates the sign map for the quantized transform block 604. In the sign map, negative quantized transform coefficients can be indicated with a −1 and positive quantized transform coefficients can be indicated with a 1.
The block 701 can be a block of the input video stream 300 of
An intraprediction mode 714 for encoding the block 701 can be determined by the mode decision 702. The intra prediction 704 can predict the block 701 using the intraprediction mode 714 determined by the mode decision 702. The encoder 700 generates a residual block 706, as described with respect to
The intraprediction mode 714, determined by the mode decision 702, is received by the entropy encoder 716. The entropy encoder 716 produces an entropy coded intraprediction mode. The quantized transform coefficients 710 are received by the entropy encoder 712. The entropy encoder 712 produces entropy encoded quantized coefficients. The entropy coded intraprediction mode and the entropy encoded quantized coefficients can be multiplexed (e.g., added) into the bitstream 720 by the multiplexer 718.
The bitstream 720 can contain a current block of the input video to be reconstructed (i.e., decoded). The bitstream 720 can be the compressed bitstream 420 of
The demultiplexer 752 receives and extracts from the bitstream 720 the entropy coded intraprediction mode and the entropy coded quantized transform coefficients. The entropy decoder 756 decodes the entropy coded quantized transform coefficients to produce quantized transform coefficients 758. The inverse quantization/transform 760 uses the quantized transform coefficients 758 to produce residual block 762. In parallel (i.e., separately), the entropy decoder 754 decodes an intraprediction mode 766 from the entropy coded intraprediction mode. The intraprediction mode 766 can be used by the intra prediction 768 to generate a prediction block. The decoder 750 uses the prediction block and the residual block 762 to generate the reconstructed block 764.
The encoder 800 may include a number of intraprediction modes. For example, the encoder 800 (and any other encoder according to implementations of this disclosure) can include 35 intraprediction modes. For example, the intraprediction modes can include a planar intraprediction mode (also referred to as mode 0), a DC intraprediction mode (also referred as mode 1), and angular prediction modes (e.g., mode 2 to mode 34).
The intraprediction modes can be classified into classes. The classes can include: horizontal, vertical, and diagonal or other. Horizontal intraprediction modes are directional prediction modes (e.g., modes 614) whose intraprediction angles are close to the horizontal axis (i.e., are generally horizontal). Vertical intraprediction modes are directional intraprediction modes (e.g., modes 2230) whose prediction angles are close to the vertical axis (i.e., are generally vertical). Diagonal or other intraprediction modes are all the remaining directional (which are neither horizontal nor vertical) and nondirectional intraprediction modes. The diagonal or other intraprediction modes are the intraprediction modes 05, 1521, and 3134. “Diagonal or other” may be referred to herein, collectively, as “diagonal.”
A description is now given as to how the encoder 800 can use the intraprediction mode to encode the quantized coefficients. Let m denote the intraprediction mode 714 used to generate the residual block 706, and let c(m) denote the class into which the intraprediction mode m is classified. For example, m can have a value equal to, or corresponding to, one of the intraprediction modes 0 to 34. Depending on the intraprediction angle, the classification c(m) of the intraprediction mode m can be one of horizontal, vertical, and diagonal or other. For example, if the prediction angle is more horizontal than vertical, then the intraprediction mode can be classified as horizontal (i.e., c(m)=horizontal). If the prediction angle is more vertical than horizontal or diagonal, then the intraprediction mode can be classified as vertical (i.e., c(m)=vertical). Otherwise, the intraprediction mode can be classified as diagonal (i.e., c(m)=diagonal). A classification of “diagonal” includes diagonal and other.
The encoder 800 can use the classification c(m) to derive one or more context models to be used for encoding the nonzero map, such as the nonzero map 606 of
The bitstream 720 can contain a current block of the input video to be reconstructed (i.e., decoded). The bitstream 720 can be the compressed bitstream 420 of
The demultiplexer 752 receives the bitstream 720. The demultiplexer extracts, from the bitstream 720, the entropy coded intraprediction mode and the entropy coded quantized transform coefficients. The entropy decoder 754 decodes the entropy coded intraprediction mode to generate the intraprediction mode 766. The entropy decoder 855 receives the intraprediction mode 766, along path 852, and receives the entropy coded quantized transform coefficients to generate the quantized transform coefficients 758. That is, the entropy decoder 855 can use the intraprediction mode to determine a context model for decoding the quantized coefficients of the transform block. The inverse quantization/transform 760 uses the quantized transform coefficients 758 to produce residual block 762. The intraprediction mode 766 can be used by the intra prediction stage 508 to generate a prediction block. The decoder 750 uses the prediction block and the residual block 762 to generate the reconstructed block 764.
The process 900 can be implemented using specialized hardware or firmware. Some computing devices can have multiple memories, multiple processors, or both. The steps or operations of the process 900 can be distributed using different processors, memories, or both. Use of the terms “processor” or “memory” in the singular encompasses computing devices that have one processor or one memory as well as devices that have multiple processors or multiple memories that can be used in the performance of some or all of the recited steps.
The process 900 can be implemented in an encoder such as the encoder 400 of
At 902, the process 900 receives a block of quantized transform coefficients. When implemented in an encoder, the process 900 can receive the block of quantized transform coefficients generated as described above with respect to
At 904, the process 900 determines a context for coding the intraprediction mode using the quantized transform coefficients. When implemented by an encoder, the process 900 determines the context (i.e., context model) for encoding the intraprediction mode in the encoded bitstream. When implemented by a decoder, the process 900 determines the context for decoding the intraprediction mode from the encoded bitstream.
Encoding the intraprediction mode involves binarizing (i.e., converting to a binary code) the magnitude of a value corresponding to the intraprediction before encoding the binarized value using the context model. The context model provides estimates of conditional probabilities for coding the binary symbols of a binarized intraprediction mode. Improving the coding context can result in using less bits to encode (i.e., entropy encode) the possible values of intraprediction modes.
Implementations according to this disclosure leverage principles of information theory to improve the coding of intraprediction modes. Let X and Y denote two random variables with finite alphabets (i.e., possible values). Using the chain rule, the joint entropy H(X,Y) of X and Y can be decomposed as given by equation (1):
H(X,Y)=X(X)+H(YX)=H(Y)+H(XY) (1)
In equation (1), H(X) and H(Y) denote the marginal entropy of X and Y, respectively, H(YX) denotes the conditional entropy of Y given X, and H(XY) denotes the conditional entropy of X given Y. If X denotes the residual block and Y denotes the intraprediction mode associated with the residual block X, and in the case where the residual block Y and the intraprediction mode X are correlated, then it follows from the chain rule that H(YX)<H(Y). That is, encoding the intraprediction mode (i.e., Y) using information in the residual block (i.e., H(YX)) results in more compression than directly encoding the intraprediction mode (i.e., H(Y)). As such, compression performance can be improved by leveraging the information of the residual block (i.e., X) in coding the intraprediction mode (i.e., Y).
Using information in the residual block includes using information in the residual block itself, using information in the corresponding transform block, or using information in the corresponding quantized transform block. If I(X;Y) denotes the mutual information between X (i.e., the residual block) and Y (i.e., the intraprediction mode), equation (1) results in I(X;Y)=H(Y)−H(YX)>0.
The process 900 can leverage the correlation between the residual block and the intraprediction mode for encoding and decoding the intraprediction mode. For example, the correlation can be expressed as: if the prediction angle is close to the horizontal axis (i.e., is generally horizontal), the nonzero quantized transform coefficients are more likely to appear in the first several columns of the quantized transform block; and if the prediction angle is close to the vertical axis (i.e., is generally vertical), the quantized transform coefficients are more likely to appear in the first several rows of the quantized transform block.
In an example, determining the coding context can include determining a category for the intraprediction mode using the quantized transform coefficients of the quantized transform block. The category can have a value that is selected from a set that includes a first category value and a second category value. The set can include other values. In an example, determining the coding context can be based on a comparison of a first number of nonzero coefficients in a row of the quantized transform block to a second number of nonzero coefficient in a column of the quantized transform block. The determination can be made using a nonzero map, such as the nonzero map 606 of
In an example, determining a category for the intraprediction mode includes determining a difference between a first number of nonzero coefficients in a row of the quantized transform block and a second number of nonzero coefficient in a column of the quantized transform block. The row can be a first row (e.g., topmost row) of the quantized transform block and the column can be a first column (e.g., leftmost column) of the quantized transform block. The first row can include more than the topmost row and the first column can include more than the leftmost column. For example, the number of rows and columns can depend on the size of the block. For example, for an 8×8 block, the first row can include the two topmost rows and the first column can include the two leftmost columns, and for a 32×32 block, the first row can include the four topmost rows and the first column can include the four leftmost columns.
Determining the category for the intraprediction mode can also include, on condition that the difference is greater than a first threshold, setting the category to a first category value and, on condition that the difference is less than a second threshold, setting the category to a second category value. That is, the category can be set to one or a first category value and a second category value based on the comparison. Determining the category for the intraprediction mode can also include, on condition that the difference is not greater than the first threshold and is not less than the second threshold, setting the category to a third category. The first category, the second category, and the third category can indicate, respectively, that the intraprediction mode is generally horizontal, generally vertical, and neither horizontal nor vertical (i.e., diagonal).
For example, let r[i] denote the number of nonzero coefficients in the ith row of the quantized transform block, and let c[j] denote the number of nonzero coefficients in the jth column of the quantized transform block, where i=0, 1, . . . , N−1 and j=0, 1, . . . , N−1 for a N×N block. In
The process 900 can compare the number of nonzero coefficients in the first row (i.e., r[0]=3) to the number of nonzero coefficients in the first column (i.e., c[0]=3). The number of nonzero coefficients can be compared to determine a category t of the intraprediction mode. In an example, the category t can have the values 1, 2, and 0 corresponding, respectively, to a horizontal intraprediction mode, a vertical intraprediction mode, and neither horizontal nor vertical intraprediction mode (i.e., diagonal). The compression gain of an encoder can be related to the number of possible values of the category t. If the category t can have three (3) values (e.g., 0, 1, 2), then the compression gain in encoding the intraprediction mode can be log_{2 }(3) (i.e., 2) bits for each block that is encoded using intra prediction.
In one example, the process 900 can determine the category t using algorithm (2):
That is, if the number of nonzero quantized transform coefficients in the first column (i.e., c[0]) is greater than the number of nonzero quantized transform coefficients in the first row (i.e., r[0]) by at least a first threshold, then the intraprediction mode can be categorized as a horizontal intraprediction mode (i.e., t=1). If the number of nonzero quantized transform coefficients in the first row (i.e., r[0]) is greater than the number of nonzero quantized transform coefficients in the first column (i.e., c[0]) by at least the second threshold, then the intraprediction mode can be categorized as a vertical intraprediction mode (i.e., t=2). Otherwise, the intraprediction mode can be categorized as diagonal. As described above, the planar intraprediction mode and the DC intraprediction mode can be categorized as “diagonal.” A category of t=0 (i.e., diagonal) can indicate that the nonzero quantized coefficients are spread throughout the quantized transform block (i.e., not concentrated enough in rows or columns) and, as such, the intraprediction mode cannot be determined to be horizontal or vertical.
In an example, the first threshold and/or the second threshold can relate to the size of the block of quantized transform coefficients. For example, for a quantized transform block of size N×N, the first threshold can be set to N/4 and the second threshold can be set to N/4. The first threshold and the second threshold can have other values that may, but need not, be related to the quantized transform block size.
Applying the algorithm (2) to the quantized transform block 604 of
In an example, if the category of intraprediction mode cannot be determined to be either horizontal or vertical based on a comparison of the number of nonzero quantized transform coefficients in a first row (e.g., the topmost row) of the quantized transform block to the number of nonzero quantized transform coefficients of a first column (e.g., left most column) of the quantized transform block, then a comparison can be made for additional rows and columns. For example, on condition that the difference is not greater than the first threshold and is not less than the second threshold, determining the category based on a comparison of a third number of nonzero coefficients in a second row of the quantized transform block to a fourth number of nonzero coefficient in a second column of the quantized transform block. The first row and the second row are different rows of the quantized transform block. The first column and the second column are different columns of the quantized transform block.
In an example, determining the coding context for the intraprediction mode can include determining a first coding context using previously coded intraprediction modes and determining the coding context using the first coding context and the category. The first coding context can be a coding context that is derived based on previously decoded intraprediction modes and/or previously decoded blocks. The category t can be used to refine the first coding context to determine the coding contest that can provide better compression than the first coding context. For example, if the first coding context has one value (e.g., an index of a coding context), the category t can be used to split the value into three (e.g., each value corresponding to a possible value of t) different context index values.
The process 900 can determine a first coding context, intra_mode_ctx, as described with respect to the entropy encoder 716 of
new_intra_mode_ctx=f(intra_mode_ctx, t) (3)
The function f of equation (3) can be a function that maps (intra_mode_ctx, t) into an index in an allowable range of available coding contexts. Any number of mapping functions can be available as long as there exists at least one intra_mode_ctx such that (intra_mode_ctx, t) is mapped to at least two distinct indices. In other words, for the at least one intra_mode_ctx, there exist two values of t, t_{a }and t_{b}, such that t_{a}≠t_{b }and (intra_mode_ctx, t_{a}) (intra_mode_ctx, t_{b}). An example of a mapping function f can be given by equation (4) below. Another example can be given by equation (5)
f(intra_mode_ctx, t)=t*NUM_CONTEXTS+intra_mode_ctx (4)
f(intra_mode_ctx, t)=intra_mode_ctx*3+t. (5)
In equation (4), NUM_CONTEXTS can be the number of distinct values of intra_mode_ctx (i.e., the number of context models in the set of coding contexts for the intraprediction mode). The process 900 can have available more context models to select the coding context (i.e., new_intra_mode_ctx) than are available for selecting intra_mode_ctx. As such, an encoder that implements the process 900, such as the encoder 1000 of
At 906, the process 900 codes the intraprediction mode using the coding context. When implemented by an encoder, coding the intraprediction mode using the coding context includes encoding, in an encoded bitstream, the intraprediction mode using the context. When implemented by a decoder, coding the intraprediction mode using the coding context includes decoding, from an encoded bitstream, the intraprediction mode using the context.
In some implementations, the process 900, at 904, can use a portion of a row and a portion of a column of the quantized transform block to determine the category t. The portion can be any contiguous or noncontiguous portion. In an example, the portion of the row (column) can be the second half of the row (column). That is, for a quantized transform block of size N×N where a row Row(i) (Column(i)) includes the quantized transform coefficients Row(i, x) (Column(x, i)), where x=0, 1, . . . , N−1, the second half of the row (column) includes the quantized transform coefficients Row(i, m) (Column(m, i)), where m=N/2, . . . , N−1. If r[i] denotes the number of nonzero coefficients in the second half of the ith row and let c[j] denote the number of nonzero coefficients in the second half of the jth column, then the category t can be determined using algorithm (6):
In algorithm (6), if the first column of the quantized transform block includes nonzero quantized transform coefficients in the second portion of the first column (i.e., c[0]>0) and there are no nonzero quantized transform coefficients in the second portion of the first row (i.e., r[0]==0), then the category t is set to 1 indicating, for example, a generally horizontal intraprediction mode. If the first row of the quantized transform block includes nonzero quantized transform coefficients in the second portion of the first row (i.e., r[0]>0) and there are no nonzero quantized transform coefficients in the second portion of the first column (i.e., c[0]==0), then the category t is set to 2 indicating, for example, a generally vertical intraprediction mode. Otherwise, the category t is to zero indicating, for example, neither horizontal nor vertical intraprediction mode. In some examples, instead of setting the category t to 0 after evaluating the portions of the first row and column, the algorithm (6) can be applied to another row and column. The process 900 can be applied repeatedly up to a predetermined number of rows and columns until a category of 1 or 2 can be set for the category t. For example, the process 900 can stop after N/2 rows and columns are examined.
The algorithm (6) can use thresholds other than 0 to determine the category t. For example, the category can be determined using algorithm (7):
In algorithm (7), the third threshold and the fourth threshold can be such that 0≤fourth threshold≤third threshold.
In yet another example, the process 900 can determine the category t using the values of the quantized transform coefficients. For example, the absolute values of the coefficients can be used. For example, let r[i] denote the sum of the absolute values of the quantized transform coefficients in the ith row of the quantized transform block and let c[i] denote the sum of the absolute values of the quantized transform coefficients of the transform block in the jth column. The value of category t can be determined using algorithm (8):
In algorithm (8), if the total sum of the absolute values of the quantized transform coefficients in the first column is greater than twice the total sum of the absolute values of the quantized transform coefficients in the first row, then the category t is set to 1 indicating, for example, a generally horizontal intraprediction mode. If the total sum of the absolute values of the quantized transform coefficients in the first row is greater than twice the total sum of the absolute values of the quantized transform coefficients in the first column, then the category t is set to 2 indicating, for example, a generally vertical intraprediction mode. Otherwise, the category t is set to 0.
While the category t is described as having the values 0, 1, and 2, implementations according to this disclosure are no so limited. The category t can have any number of values. For example, additional values of the category t can be used to provide additional granularity regarding the angle of the horizontal and vertical intraprediction modes.
In an implementation, the process 900 can include determining an intraprediction mode predictor p using the quantized transform coefficients and encoding whether the intraprediction mode predictor and the intra prediction mode are the same. The intraprediction mode predictor p inversely predicts the intraprediction mode based on the distribution of the decoded quantized transform coefficients. If, for example, the decoded nonzero quantized transform coefficients are skewed toward the first several columns of the quantized transform block, then a generally horizontal intraprediction mode can be selected. If the decoded nonzero quantized transform coefficients are skewed toward the first several rows of the quantized transform block, then a generally vertical intraprediction mode can be selected.
The intraprediction mode predictor p can be determined using a significance map of the quantized transform coefficients. The significance map can be as described with respect to the nonzero map 606 of
The process 900 can determine the intraprediction mode predictor p based on a comparison of a number of nonzero quantized coefficients in rows and columns of the quantized transform block as described above. The intraprediction mode predictor p can be indicative of the prediction angle. In an example, the intraprediction mode predictor p can have the value of an intraprediction mode that is generally horizontal or generally vertical. For example, and using the intraprediction modes described with respect to
The process 900, when implemented by an encoder, can set a syntax element to a value (e.g., 0) when the intraprediction mode predictor and the intra prediction mode are the same. If the indicators indicates that the intraprediction mode m is equal to the intraprediction mode predictor p (i.e., m=p), then a decoder can use the encoded intraprediction mode to decode the current block. The syntax element can be set another value (e.g., 1), followed by an encoding of the intraprediction mode, when the intraprediction mode predictor and the intra prediction mode are not the same.
As another example, the intra prediction mode m can be encoded (and decoded by a decoder such as a decoder 1200 of
As the quantized transform coefficients are available at a decoder (e.g., in an encoded bitstream), the decoder can determine the category t and the intra prediction mode predictor p from the quantized transform coefficients as described above.
The encoder 1000 can use information in residual blocks to improve the compression performance of intraprediction modes. The encoder 1000 includes some of the same components as those of the encoder 700. Descriptions of components of the encoder 1000 that are common with or similar to, the encoder 700 are omitted. The encoder 1000 includes the mode decision 702, the intra prediction 704, the transform/quantization 708, the inverse quantization/transform 722, an entropy encoder 1002, the entropy encoder 712, and the multiplexer 718. In an example, the encoder 1000 can provide improvements to the encoder 700 of
The block 701 can be a block of the input video stream 300 of
An intraprediction mode 714 for encoding the block 701 can be determined by the mode decision 702. The intra prediction 704 can encode the block 701 using the intraprediction mode 714 determined by the mode decision 702. The encoder 700 generates a residual block 706, as described with respect to
The entropy encoder 712 produces entropy encoded quantized coefficients. The entropy encoder 1002 receives the intraprediction mode 714 determined by the mode decision 702 and receives, along path 1004, the quantized transform coefficients 710. The entropy encoder 1002 uses the quantized transform coefficients 710, as described with respect to
The process 1100 can be implemented using specialized hardware or firmware. Some computing devices can have multiple memories, multiple processors, or both. The steps or operations of the process 1100 can be distributed using different processors, memories, or both. Use of the terms “processor” or “memory” in the singular encompasses computing devices that have one processor or one memory as well as devices that have multiple processors or multiple memories that can be used in the performance of some or all of the recited steps.
The process 1100 can be implemented in a decoder such as the decoder 500 of
At 1102, the process 1100 decodes, from an encoded bitstream, quantized transform coefficients of a quantized transform block.
At 1104, the process 1100 determines a category for the intraprediction mode using the quantized transform coefficients. The category can be one of horizontal, vertical, and other (i.e., not horizontal or vertical, e.g., diagonal). For example, the process 1100 can determine a distribution of the nonzero quantized transform coefficients, as described above with respect to the process 900, to determine the category of the intraprediction mode. For example, if the nonzero coefficients of the quantized transform block are concentrated in the first several rows (e.g., the first row), then the intraprediction mode is determined to be likely vertical; if the nonzero coefficients are concentrated in the first several columns (e.g., the first column), then the intraprediction mode is determined to be likely horizontal. If the nonzero coefficients are not concentrated in rows or columns, then the intraprediction mode is more likely to be diagonal (e.g., DC, planar, or a diagonal prediction mode). As such, depending on the distribution of the nonzero decoded quantized coefficients, the intraprediction mode can be classified as described above with respect to the category t (e.g., into one three values).
In an example, decoding the current block using the category includes determining a first context for decoding the intraprediction mode, determining a second context model using the category and the first context, and decoding the intraprediction mode using the second context. The first context and the second context can be determined as described above with the respect to
At 1106, the process 1100 decodes, from the encoded bitstream, the current block using the category.
In some implementations, the process 1100 can include decoding, from the encoded bitstream, whether an intraprediction mode predictor is the same as the intraprediction mode. The intraprediction mode predictor can be a prediction of the intraprediction mode and is determined by an encoder using the quantized transform coefficients as described with respect to
The decoder 1200 can use quantized transform coefficient information in an encoded bitstream to determine an intra prediction mode for decoding the current block. The encoded bitstream can be the bitstream 720 that is received from, generated by, or output by an encoder such as the encoder 1000 of
The decoder 1200 includes some of the same components as those of the decoder 750 of
The entropy decoder 1202 receives, along path 1204, the quantized transform coefficients 758, which are entropy decoded by the entropy decoder 756. The entropy decoder 1202 also receives an entropy encoded intraprediction mode from the demultiplexer 752. The entropy decoder 1202 can determine a category for the intraprediction mode as described with respect to
At 1302, the process 1300 decodes a block of quantized transform coefficients from the encoded bitstream. Decoding the block of quantized transform coefficients can be as described with respect to the process 900 at 902. At 1304, the process 1300 determines at least one context for the intraprediction mode using the quantized transform coefficients. Determining the context can be as described above with respect to the process 900 at 904. The encoded bitstream can include more than one intraprediction modes for the current block. As such, at least one intraprediction mode can be determined using the quantized transform coefficients.
At 1306, the process 1300 decodes the intraprediction mode from the encoded bitstream using the at least one context. At 1308, the process 1300 reconstructs a residual block using the block of quantized transform coefficients. When the process 1300 is implemented in an encoder, reconstructing the residual block can be as described with respect to the intra prediction 704 of
At 1310, the process 1300 reconstructs the current block using the reconstructed residual block and the intraprediction mode. When the process 1300 is implemented in an encoder, reconstructing the current block can be as described with respect to the reconstruction path of
In some implementations (not shown), instead of using the zigzag scan order for the first group, other scan orders may be used. For example, an alternating scan order S1 as follows may be used to map a scan position k to a block position (e.g., Cartesian position).
S1[k]=(0, k>>1) if k is even
S1[k]=((k+1)>>1, 0) if k is odd
where k=0, . . . , 2*N−2. In some implementations (not shown), a 225degree diagonal scan order (downward to the left) may be used for a group of elements from the block.
In some implementations, multilevel scan orders may be used. In these cases, each element in
When multilevel scan orders are used, once the scan order is determined for subblocks, the scan order within a subblock may be derived similarly according to the example above. In some implementations, the scan order within a subblock may inherit the scan order determined at the subblock level for the group including the element representing the subblock. For example, each subblock in 1560 may use the horizontal scan order, each subblock in 1570 may use the vertical scan order, each subblock in 1580 may use the zigzag scan order, and each subblock in 1590 may use the diagonal scan order.
The process 1600 includes partitioning 1610 a block of video data into a plurality of groups of elements. The groups may be disjoint. The block may include a twodimensional array of elements (e.g., a 4×4, 8×8, 16×16, 32×32, or any other square or rectangular block size). A group from the plurality of groups of elements may be continuous or discontinuous within the structure of the block. For example, the elements of the block may be pixel values (e.g., luminosity values). For example, the elements of the block may be quantized transform coefficients. For example, the elements of the block may be quantized transform coefficients of a residual block (e.g., a residual block resulting from intraframe prediction or a residual block resulting from interframe prediction). For example, the elements of the block may be based on a result of interframe prediction. For example, the block may be a residual block resulting from interframe prediction. For example, the elements of the block may be based on a result of intraframe prediction. For example, the block may be a residual block resulting from intraframe prediction. In some implementations, the plurality of groups includes two groups (e.g., as shown in the example of
The process 1600 includes encoding 1620, using an entropy encoder using a first scan order, elements of a first group from the plurality of groups of elements. The first scan order may be a default or initial scan order that a compatible entropy decoder is configured to use to decode elements of the first group for the block. For example, the first scan order may be a zigzag order (e.g., as shown in
The process 1600 includes selecting 1622 a next group (e.g., a second group) from the plurality of groups of elements of the block.
The process 1600 includes determining 1630, based on the elements of one or more prior groups (e.g., the first group), a next scan order (e.g., a second scan order) for the next group (e.g., the second group) from the plurality of groups of elements. In some implementations, the next scan order may be determined 1630 by selecting from among options including horizontal, vertical, diagonal, zigzag, etc. (e.g., as illustrated in
The process 1600 includes encoding 1640, using the entropy encoder using the next scan order (e.g., the second scan order), elements of the next group (e.g., the second group) from the plurality of groups of elements of the block. For example, the next scan order may be a horizontal scan order, a vertical scan order, a zigzag scan order, etc. (e.g., as shown in
If (at 1645) the last group from the plurality of groups of elements of the block has not yet been encoded, then the next group is selected 1622 and the processing of groups continues to encode remaining elements of the block.
When (at 1645) the last group from the plurality of groups of elements of the block has been encoded, the encoded elements of the block are output 1650 to an encoded bitstream. For example, the encoded elements may be output from the entropy encoder 712 through the multiplexer 718 to be included in the bitstream 720 that bears a video signal including the block. In some implementations, the bitstream may be stored (e.g., in memory or on a disk) or transmitted (e.g., via network communication link). In some implementations (not shown), encoded elements may be output to an encoded bitstream as they become available.
In some implementations, the process 1600 adjusts the scan order for the next stage on the fly based on information collected at a current stage. For example, let X denote an N×M block that has N>0 rows and M>0 columns. The process 1600 may be implemented by specialized hardware and/or software that performs operations corresponding to the following pseudocode:
1. Partition X into K nonoverlapping groups, denoted by X[1], X[2], . . . , X[K], respectively.
2. Use a default scan order (e.g., zigzag, diagonal, etc.) to encode X[1].
3. Set k=2.
4. Determine the scan order to encode X[k] based on X[1], . . . , X[k−1].
5. Encode X[k] by using the determined scan order.
6. Increase k by 1.
7. Repeat Steps 46 until k=K+1.
The process 1700 includes partitioning 1710 a block of video data into a plurality of groups of elements. The groups may be disjoint. The block may include a twodimensional array of elements (e.g., a 4×4, 8×8, 16×16, 32×32, or any other square or rectangular block size). A group from the plurality of groups of elements may be continuous or discontinuous within the structure of the block. For example, the elements of the block may be pixel values (e.g., luminosity values). For example, the elements of the block may be quantized transform coefficients. For example, the elements of the block may be quantized transform coefficients of a residual block (e.g., a residual block resulting from intraframe prediction or a residual block resulting from interframe prediction). For example, the elements of the block may be based on a result of interframe prediction. For example, the block may be a residual block resulting from interframe prediction. For example, the elements of the block may be based on a result of intraframe prediction. For example, the block may be a residual block resulting from intraframe prediction. In some implementations, the plurality of groups includes two groups (e.g., as shown in the example of
The process 1700 includes decoding 1720, using an entropy decoder, data from an encoded bitstream to obtain, using a first scan order, elements of a first group from the plurality of groups of elements. The first scan order may be a default or initial scan order that a compatible entropy encoder is configured to use to encode elements of the first group for the block. For example, the first scan order may be a zigzag order (e.g., as shown in
The process 1700 includes selecting 1722 a next group (e.g., a second group) from the plurality of groups of elements of the block.
The process 1700 includes determining 1730, based on the elements of one or more prior groups (e.g., the first group), a next scan order (e.g., a second scan order) for the next group (e.g., the second group) from the plurality of groups of elements. In some implementations, the next scan order may be determined 1730 by selecting from among options including horizontal, vertical, diagonal, zigzag, etc. (e.g., as illustrated in
The process 1700 includes decoding 1740, using the entropy decoder, data from the encoded bitstream to obtain, using the next scan order (e.g., the second scan order), elements of next group (e.g., the second group) from the plurality of groups of elements of the block. For example, the next scan order may be a horizontal scan order, a vertical scan order, a zigzag scan order, etc. (e.g., as shown in
If (at 1745) the last group from the plurality of groups of elements of the block has not yet been decoded, then the next group is selected 1722 and the processing of groups continues to decode remaining elements of the block.
When (at 1745) the last group from the plurality of groups of elements of the block has been decoded, the decoded elements of the block are output 1750 for processing to generate video. For example, the decoded elements may be output 1750 from the entropy decoder 756 through the inverse quantization/transform 760 and used to generate the reconstructed block 764 that forms part of a video. In some implementations, the resulting video may be stored (e.g., in the memory 204 or the secondary storage 214), transmitted (e.g., via a network communication link), or displayed (e.g., in a user interface, such as on screen or a projection). For example, the process 1700 may include displaying video that is generated based in part on the decoded elements of the block (e.g., where the elements of the block are quantized transform coefficients of a residual block).
In some implementations, the process 1700 adjusts the scan order for the next stage on the fly based on information collected at a current stage. For example, let X denote an N×M block that has N>0 rows and M>0 columns. The process 1700 may be implemented by specialized hardware and/or software that performs operations corresponding to the following pseudocode:
1. Partition X into K nonoverlapping groups, denoted by X[1], X[2], . . . , X[K], respectively.
2. Use a default scan order (e.g., zigzag, diagonal, etc.) to decode X[1].
3. Set k=2.
4. Determine the scan order to decode X[k] based on X[1], . . . , X[k−1].
5. Decode X[k] by using the determined scan order.
6. Increase k by 1.
7. Repeat Steps 46 until k=K+1.
The process 1800 includes determining 1810 a first sum of magnitudes of elements in a portion of the first row of the block. For example, the portion of the first row may be a right half of the first row. In some implementations, the portion of the first row may be the whole first row. In some implementations, the elements may be constrained to take integer values (e.g., as a result of quantization). For example, the magnitude of an element may be the absolute value of the element. In some implementations, the magnitude of an element may be represented as the square of the value of the element or another quantity that varies monotonically with the absolute value of the element. For example, the first sum may be determined 1810 as:
SR=r[ceil(M/2)]+r[ceil(M/2)+1]+ . . . r[M−1] (9)
where SR is the first sum; M is the number of columns in the block (e.g., an N×M block); r[i] is the ith element of the first row of the block for i=0 to M−1; and ceil( ) denotes the ceiling function, i.e., for any real number a, ceil(a) returns the smallest integer that is greater than or equal to a. M may be greater than or equal to two. For example, the elements of the block may be quantized transform coefficients of a residual block (e.g., a residual block resulting from intraframe prediction or a residual block resulting from interframe prediction).
The process 1800 includes determining 1812 a second sum of magnitudes of elements in a portion of the first column of the block. For example, the portion of the first column may be a bottom half of the first column. In some implementations, the portion of the first row may be the whole first column. In some implementations, the elements may be constrained to take integer values (e.g., as a result of quantization). For example, the magnitude of an element may be the absolute value of the element. In some implementations, the magnitude of an element may be represented as the square of the value of the element or another quantity that varies monotonically with the absolute value of the element. For example, the second sum may be determined 1812 as:
SC=c[ceil(N/2)]+c[ceil(N/2)+1]+ . . . c[N−1] (10)
where SC is the second sum; N is the number of rows in the block (e.g., an N×M block); and c[i] is the ith element of the first column of the block for i=0 to N−1. N may be greater than or equal to two.
The process 1800 may include, responsive to the first sum (e.g., SR of equation 9) being zero and the second sum (e.g., SC of equation 10) being positive (at 1815), determining 1820 the next scan order (e.g., a second scan order) to be a vertical scan order (e.g., as illustrated by the scan order 1520 of
The process 1800 may include, responsive to the first sum (e.g., SR of equation 9) being positive and the second sum (e.g., SC of equation 10) being zero (at 1825), determining 1830 the next scan order (e.g., the second scan order) to be a horizontal scan order (e.g., as illustrated by the scan order 1510 of
Otherwise, the process 1800 may include, responsive to the first sum (e.g., SR of equation 9) and the second sum (e.g., SC of equation 10) both being positive or both being zero, determining 1832 the next scan order (e.g., the second scan order) to be a default scan order (e.g., zigzag, diagonal, etc.).
In some embodiments (not shown), SR==0 at 1815 may be replaced by SR<=fifth threshold and SC>0 may be replaced by SC>=sixth threshold. Similarly, SR>0 at 1825 may be replaced by SR>=sixth threshold and SC==0 at 1825 may be replaced by SC<=fifth threshold.
The process 1840 includes determining 1850 a first count of nonzero elements in a portion of the first row of the block. For example, the portion of the first row may be a right half of the first row. In some implementations, the portion of the first row may be the whole first row. In some implementations, the first count may be determined 1850 based on a nonzero map (e.g., similar to the nonzero map 606) for available group(s) of the block. For example, the first count may be determined 1850 as a sum of the binary values from a nonzero map corresponding to the portion of the first row of the block. For example, the first count may be determined 1850 as:
CR=nz(r[ceil(M/2)])+nz(r[ceil(M/2)+1])+ . . . nz(r[M−1]) (11)
where CR is the first count; M is the number of columns in the block (e.g., an N×M block); r[i] is the ith element of the first row of the block for i=0 to M−1; nz( ) denotes a nonzero function that returns one if the argument is nonzero and zero if the argument is zero; and ceil( ) denotes the ceiling function, i.e., for any real number a, ceil(a) returns the smallest integer that is greater than or equal to a. M may be greater than or equal to two. For example, the elements of the block may be quantized transform coefficients of a residual block (e.g., a residual block resulting from intraframe prediction or a residual block resulting from interframe prediction).
The process 1840 includes determining 1852 a second count of nonzero elements in a portion of the first column of the block. For example, the portion of the first column may be a bottom half of the first column. In some implementations, the portion of the first column may be the whole first column. In some implementations, the second count may be determined 1852 based on a nonzero map (e.g., similar to the nonzero map 606) for available group(s) of the block. For example, the second count may be determined 1850 as a sum of the binary values from a nonzero map corresponding to the portion of the first column of the block. For example, the second count may be determined 1852 as:
CC=nz(c[ceil(N/2)])+nz(c[ceil(N/2)+1])+ . . . nz(c[N−1]) (12)
where CC is the second count; N is the number of rows in the block (e.g., an N×M block); and c[i] is the ith element of the first column of the block for i=0 to N−1. N may be greater than or equal to two. For example, the elements of the block may be quantized transform coefficients of a residual block (e.g., a residual block resulting from intraframe prediction or a residual block resulting from interframe prediction).
The process 1840 may include, responsive to the first count (e.g., CR of equation 11) being zero and the second count (e.g., CC of equation 12) being positive (at 1855), determining 1860 the next scan order (e.g., a second scan order) to be a vertical scan order (e.g., as illustrated by the scan order 1520 of
The process 1840 may include, responsive to the first count (e.g., CR of equation 11) being positive and the second count (e.g., CC of equation 12) being zero (at 1865), determining 1870 the next scan order (e.g., the second scan order) to be a horizontal scan order (e.g., as illustrated by the scan order 1510 of
Otherwise, the process 1840 may include, responsive to the first count (e.g., CR of equation 11) and the second count (e.g., CC of equation 12) both being positive or both being zero, determining 1872 the next scan order (e.g., the second scan order) to be a default scan order (e.g., zigzag, diagonal, etc.).
In some implementations, CR==0 at 1855 may be replaced by CR<=seventh threshold and CC>0 may be replaced by CC>=eighth threshold. Similarly, CR>0 at 1865 may be replaced by CR>=eighth threshold and CC==0 at 1865 may be replaced by CC<=seventh threshold.
The process 1600 may impact context derivation in coding quantized transform coefficients. For example, depending upon the scan order selected, the available neighboring coefficients available for deriving contexts to code the current coefficient may be different. For example some entropy coding schemes may use two previously coded neighbors to derive a context for coding (encoding and decoding) a current coefficient represented by a token. Suppose that the first group in the process 1600 consists of the first row and the first column of the current block. Then the block positions of the two neighbors used for context derivation in coding a token at block position [r, c] may be given in the table below.
In some implementations, contexts used in coding coefficients in a group other than the first group may depend upon the scan order selected for that group. For example, one context may be used for coding the coefficient at position (r, c), where r>0 and c>0, in the second group of
The process 1600 may be used in tandem with processes (e.g., the process 900) for encoding an intraprediction mode for a block using context determined based on elements of the block. In some implementations, such a combination of processes (e.g., of the process 1600 with the process 900) may be used to more fully leverage a correlation between intraprediction modes and quantized transform coefficients in entropy coding.
The process 1900 includes partitioning 1910 a block of video data into a plurality of groups of elements. The groups may be disjoint. The block may include a twodimensional array of elements (e.g., a 4×4, 8×8, 16×16, 32×32, or any other square or rectangular block size). A group from the plurality of groups of elements may be continuous or discontinuous within the structure of the block. For example, the elements of the block may be pixel values (e.g., luminosity values). For example, the elements of the block may be quantized transform coefficients. For example, the elements of the block may be quantized transform coefficients of a residual block (e.g., a residual block resulting from intraframe prediction or a residual block resulting from interframe prediction). For example, the elements of the block may be based on a result of interframe prediction. For example, the block may be a residual block resulting from interframe prediction. For example, the elements of the block may be based on a result of intraframe prediction. For example, the block may be a residual block resulting from intraframe prediction. In some implementations, the plurality of groups includes two groups (e.g., as shown in the example of
The process 1900 includes encoding 1920, using an entropy encoder using a first scan order, elements (e.g., quantized transform coefficients of an intraprediction residual block) of a first group from the plurality of groups of elements. The first scan order may be a default or initial scan order that a compatible entropy decoder is configured to use to decode elements of the first group for the block. For example, the first scan order may be a zigzag order (e.g., as shown in
The process 1900 includes selecting 1922 a next group (e.g., a second group) from the plurality of groups of elements of the block. For example, the next group may be selected 1922 by incrementing a group index.
The process 1900 includes determining 1930, based on the elements of one or more prior groups (e.g., the first group), a next scan order (e.g., a second scan order) for the next group (e.g., the second group) from the plurality of groups of elements. In some implementations, the next scan order may be determined 1930 by selecting from among options including horizontal, vertical, diagonal, zigzag, etc. (e.g., as illustrated in
The process 1900 includes encoding 1940, using the entropy encoder using the next scan order (e.g., the second scan order), elements of the next group (e.g., the second group) from the plurality of groups of elements of the block. For example, the next scan order may be a horizontal scan order, a vertical scan order, a zigzag scan order, etc. (e.g., as shown in
If (at 1945) the last group from the plurality of groups of elements of the block has not yet been encoded, then the next group is selected 1922 and the processing of groups continues to encode 1940 remaining elements of the block.
When (at 1945) the last group from the plurality of groups of elements of the block has been encoded, a category for the block is determined 1950 based on the elements (e.g., quantized transform coefficients of an intraprediction residual block) of the block. For example, the category may be determined as one of horizontal, vertical, and other (i.e., not horizontal or vertical, e.g., diagonal). The category may indicate a directional pattern in a distribution of nonzero elements within the block, which may be related to an intraprediction mode that was used to generate the block. In some implementations, the category may be determined 1950 based on a first count of nonzero elements in a first portion of the block (e.g., a right half of the block) and based on a second count of nonzero elements in a second portion of the block (e.g., a bottom half of the block). For example, if the first count is zero and the second count is positive, the category may be determined 1950 to be vertical. For example, if the first count is positive and the second count is zero, the category may be determined 1950 to be horizontal. For example, if the first count and the second count are both zero or both positive, the category may be determined 1950 to be diagonal. In some implementations, the category may be determined 1950 based on elements in the first group of elements of the block. For example, the category may be determined 1950 to be vertical when CR of equation 11 is zero and CC of equation 12 is positive. For example, the category may be determined 1950 to be horizontal when CR of equation 11 is positive and CC of equation 12 is zero. For example, the category may be determined 1950 to be diagonal when CR of equation 11 and CC of equation 12 are both zero or both positive. In some implementations, the category (t) may be determined using the algorithm (2) described in relation to
The process 1900 includes determining 1952 a context based on the category. Because patterns in the distribution of nonzero elements of the block may be correlated with an intraprediction mode that was used to generate the block, the category may be used to select a more useful context for the intraprediction mode, which may improve coding efficiency for encoding the intraprediction mode. For example, the context may be determined 1952 as described with respect to operation 904 of
The process 1900 includes encoding 1954, using an entropy decoder with the context, an intraprediction mode for the block. For example, the intraprediction mode may be encoded 1954 as described in relation to operation 906 of
The process 1900 includes outputting 1960 the encoded elements of the block and the encoded intraprediction mode for the block to an encoded bitstream. For example, the encoded elements may be output 1960 from the entropy encoder 712 through the multiplexer 718 to be included in the bitstream 720 that bears a video signal including the block. For example, the encoded intraprediction mode may be output 1960 from the entropy encoder 716 through the multiplexer 718 to be included in the bitstream 720 that bears a video signal including the block. In some implementations, the bitstream may be stored (e.g., in memory or on a disk) or transmitted (e.g., via network communication link). In some implementations (not shown), encoded elements may be output to an encoded bitstream as they become available.
In some implementations, the process 1900 may be implemented by specialized hardware and/or software that performs operations corresponding to the following pseudocode (which uses the definitions provided in relation to equation 11 and equation 12 above):
1. Determine an intraprediction mode m to generate prediction P for the current block X.
2. Compute the residual block E=X−P, and obtain quantized coefficients by performing transform (if applicable) and quantization.
3. Use a default scan order (e.g. zigzag, diagonal, etc.) to encode r[0], . . . , r[N−1] and c[0], . . . , c[N−1]. (An example of such a scan order is shown in
 a. if CC>=1 and CR==0, set t=1, and select the vertical scan order.
 b. if CC==0 and CR>=1, set t=2, and select the horizontal scan order.
 c. otherwise, set t=0, and select the default scan order (e.g., zigzag, diagonal, etc.).
4. Encode the intraprediction mode m by using the category t.
5. Encode the remaining coefficients of the block by using the scan order selected in Step 3.
The process 1700 may be used in tandem with techniques (e.g., the process 1100) for decoding an intraprediction mode for a block using context determined based on elements of the block. In some implementations, such a combination of techniques (e.g., of the process 1700 with the process 1100) may be used to more fully leverage a correlation between intraprediction modes and quantized transform coefficients in entropy coding.
The process 2000 includes partitioning 2010 a block of video data into a plurality of groups of elements. The groups may be disjoint. The block may include a twodimensional array of elements (e.g., a 4×4, 8×8, 16×16, 32×32, or any other square or rectangular block size). A group from the plurality of groups of elements may be continuous or discontinuous within the structure of the block. For example, the elements of the block may be pixel values (e.g., luminosity values). For example, the elements of the block may be quantized transform coefficients. For example, the elements of the block may be quantized transform coefficients of a residual block (e.g., a residual block resulting from intraframe prediction or a residual block resulting from interframe prediction). For example, the elements of the block may be based on a result of interframe prediction. For example, the block may be a residual block resulting from interframe prediction. For example, the elements of the block may be based on a result of intraframe prediction. For example, the block may be a residual block resulting from intraframe prediction. In some implementations, the plurality of groups includes two groups (e.g., as shown in the example of
The process 2000 includes decoding 2020, using an entropy decoder, data from an encoded bitstream to obtain, using a first scan order, elements of a first group from the plurality of groups of elements. The first scan order may be a default or initial scan order that a compatible entropy encoder is configured to use to encode elements of the first group for the block. For example, the first scan order may be a zigzag order (e.g., as shown in
The process 2000 includes selecting 2022 a next group (e.g., a second group) from the plurality of groups of elements of the block. For example, the next group may be selected 2022 by incrementing a group index.
The process 2000 includes determining 2030, based on the elements of one or more prior groups (e.g., the first group), a next scan order (e.g., a second scan order) for the next group (e.g., the second group) from the plurality of groups of elements. In some implementations, the next scan order may be determined 2030 by selecting from among options including horizontal, vertical, diagonal, zigzag, etc. (e.g., as illustrated in
The process 2000 includes decoding 2040, using the entropy decoder, data from the encoded bitstream to obtain, using the next scan order (e.g., the second scan order), elements of next group (e.g., the second group) from the plurality of groups of elements of the block. For example, the next scan order may be a horizontal scan order, a vertical scan order, a zigzag scan order, etc. (e.g., as shown in
If (at 2045) the last group from the plurality of groups of elements of the block has not yet been decoded, then the next group is selected 2022 and the processing of groups continues to decode 2040 remaining elements of the block.
When (at 2045) the last group from the plurality of groups of elements of the block has been decoded, a category for the block is determined 2050 based on the decoded elements (e.g., quantized transform coefficients of an intraprediction residual block) of the block. For example, the category may be determined as one of horizontal, vertical, and other (i.e., not horizontal or vertical, e.g., diagonal). The category may indicate a directional pattern in a distribution of nonzero elements within the block, which may be related to an intraprediction mode that was used to generate the block. In some implementations, the category may be determined 2050 based on a first count of nonzero elements in a first portion of the block (e.g., a right half of the block) and based on a second count of nonzero elements in a second portion of the block (e.g., a bottom half of the block). For example, if the first count is zero and the second count is positive, the category may be determined 2050 to be vertical. For example, if the first count is positive and the second count is zero, the category may be determined 2050 to be horizontal. For example, if the first count and the second count are both zero or both positive, the category may be determined 2050 to be diagonal. In some implementations, the category may be determined 2050 based on elements in the first group of elements of the block. For example, the category may be determined 2050 to be vertical when CR of equation 11 is zero and CC of equation 12 is positive. For example, the category may be determined 2050 to be horizontal when CR of equation 11 is positive and CC of equation 12 is zero. For example, the category may be determined 2050 to be diagonal when CR of equation 11 and CC of equation 12 are both zero or both positive. In some implementations, the category (t) may be determined using the algorithm (2) described in relation to
The process 2000 includes determining 2052 a context based on the category. Because patterns in the distribution of nonzero elements of the block may be correlated with an intraprediction mode that was used to generate the block, the category may be used to select a more useful context for the intraprediction mode, which may improve coding efficiency for decoding the intraprediction mode. For example, the context may be determined 2052 as described with respect to operation 904 of
The process 2000 includes decoding 2054, using an entropy decoder with the context, data from the encoded bitstream to obtain an intraprediction mode for the block. For example, the intraprediction mode may be decoded 2054 as described in relation to operation 1106 of
The process 2000 includes outputting 2060 the decoded elements of the block and the decoded intraprediction mode for the block for processing to generate video. For example, the decoded elements may be output 2060 from the entropy decoder 756 through the inverse quantization/transform 760 and used to generate the reconstructed block 764 that forms part of a video. For example, the decoded intraprediction mode may be output 2060 from the entropy decoder 754 to the intra prediction 768 and used to generate the reconstructed block 764 that forms part of a video. In some implementations, the resulting video may be stored (e.g., in memory or on a disk), transmitted (e.g., via network communication link), or displayed (e.g., in a user interface, such as on screen or a projection). For example, the process 2000 may include displaying video that is generated based in part on the decoded elements of the block (e.g., where the elements of the block are quantized transform coefficients of a residual block) and based in part on the decoded intraprediction mode.
In some implementations, the process 2000 may be implemented by specialized hardware and/or software that performs operations corresponding to the following pseudocode (which uses the definitions provided in relation to equation 11 and equation 12 above):
1. Use a default scan order (e.g. zigzag, diagonal, etc.) to decode r[0], . . . , r[N−1] and c[0], . . . , c[N−1]. (An example of such a scan order is shown in
 a. if CC>=1 and CR==0, set t=1, and select the vertical scan order.
 b. if CC==0 and CR>=1, set t=2, and select the horizontal scan order.
 c. otherwise, set t=0, and select the default scan order (e.g., zigzag, diagonal, etc.).
2. Decode an intraprediction mode m by using t.
3. Decode the remaining coefficients of the block by using the scan order selected in Step 1.
4. Use the intraprediction mode m to generate prediction P.
5. From the decoded coefficients, reconstruct a residual block Ê by performing inverse quantization and inverse transform (if applicable).
6. Reconstruct the current image block {circumflex over (X)}=P+Ê.
In some implementations, a category for a block, determined based on elements of the block, may be used to encode and decode an intraprediction mode m in a variety of ways, such as, deriving the context used to encode and decode the intraprediction mode m, and/or deriving a predictor for the intraprediction mode m (e.g., as described with respect to the intraprediction mode predictor p of
The aspects of encoding and decoding described above illustrate some encoding and decoding techniques. However, it is to be understood that encoding and decoding, as those terms are used in the claims, could mean compression, decompression, transformation, or any other processing or change of data.
The words “example” or “implementation” are used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as “example” or “implementation” is not necessarily to be construed as preferred or advantageous over other aspects or designs. Rather, use of the words “example” or “implementation” is intended to present concepts in a concrete fashion. As used in this application, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or”. That is, unless specified otherwise, or clear from context, “X includes A or B” is intended to mean any of the natural inclusive permutations. That is, if X includes A; X includes B; or X includes both A and B, then “X includes A or B” is satisfied under any of the foregoing instances. In addition, the articles “a” and “an” as used in this application and the appended claims should generally be construed to mean “one or more” unless specified otherwise or clear from context to be directed to a singular form. Moreover, use of the term “an implementation” or “one implementation” throughout is not intended to mean the same embodiment or implementation unless described as such.
Implementations of transmitting station 102 and/or receiving station 106 (and the algorithms, methods, instructions, etc., stored thereon and/or executed thereby, including by encoder 400 and decoder 500) can be realized in hardware, software, or any combination thereof. The hardware can include, for example, computers, intellectual property (IP) cores, applicationspecific integrated circuits (ASICs), programmable logic arrays, optical processors, programmable logic controllers, microcode, microcontrollers, servers, microprocessors, digital signal processors or any other suitable circuit. In the claims, the term “processor” should be understood as encompassing any of the foregoing hardware, either singly or in combination. The terms “signal” and “data” are used interchangeably. Further, portions of transmitting station 102 and receiving station 106 do not necessarily have to be implemented in the same manner.
Further, in one aspect, for example, transmitting station 102 or receiving station 106 can be implemented using a computer or processor with a computer program that, when executed, carries out any of the respective methods, algorithms and/or instructions described herein. In addition, or alternatively, for example, a special purpose computer/processor can be utilized which can contain other hardware for carrying out any of the methods, algorithms, or instructions described herein.
Transmitting station 102 and receiving station 106 can, for example, be implemented on computers in a video conferencing system. Alternatively, transmitting station 102 can be implemented on a server and receiving station 106 can be implemented on a device separate from the server, such as a handheld communications device. In this instance, transmitting station 102 can encode content using an encoder 400 into an encoded video signal and transmit the encoded video signal to the communications device. In turn, the communications device can then decode the encoded video signal using a decoder 500. Alternatively, the communications device can decode content stored locally on the communications device, for example, content that was not transmitted by transmitting station 102. Other transmitting station 102 and receiving station 106 implementation schemes are available. For example, receiving station 106 can be a generally stationary personal computer rather than a portable communications device and/or a device including an encoder 400 may also include a decoder 500.
Further, all or a portion of implementations of the present disclosure can take the form of a computer program product accessible from, for example, a tangible computerusable or computerreadable medium. A computerusable or computerreadable medium can be any device that can, for example, tangibly contain, store, communicate, or transport the program for use by or in connection with any processor. The medium can be, for example, an electronic, magnetic, optical, electromagnetic, or a semiconductor device. Other suitable mediums are also available.
The abovedescribed embodiments, implementations and aspects have been described in order to allow easy understanding of the present disclosure and do not limit the present disclosure. On the contrary, the disclosure is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims, which scope is to be accorded the broadest interpretation so as to encompass all such modifications and equivalent structure as is permitted under the law.