Device and method for variable length coding of video signals depending on the characteristics

• US 6,046,774 A
• Filed: 03/11/1996
• Issued: 04/04/2000
• Est. Priority Date: 06/02/1993
• Status: Expired due to Fees

• Current Assignee: LG Electronics, Inc. (LG Corporation)
• Original Assignee: Goldstar Jewellery LLC
• Inventors: Kim, Jong Kyu, Lee, In Seop, Heo, Seo Weon
• Primary Examiner(s): Chin, Tommy P.
• Assistant Examiner(s): RAO, ANAND SHASHIKANT
• Application Number:
  US08/613,547
• Time in Patent Office: 1,485 Days
• Field of Search: 348/420, 348/411, 348/416, 348/405, 348/404, 348/699, 348/700
• US Class Current: 375/240.16
• CPC Class Codes: H04N 19/114 Adapting the group of pictures [GOP] structure, e.g. number of B-frames between two anchor frames H04N19/107 takes precedence
• CPC Class Codes: H04N 19/137 Motion inside a coding unit, e.g. average field, frame or block difference
• CPC Class Codes: H04N 19/142 Detection of scene cut or scene change
• CPC Class Codes: H04N 19/149 by estimating the code amount by means of a model, e.g. mathematical model or statistical model
• CPC Class Codes: H04N 19/177 the unit being a group of pictures [GOP]
• CPC Class Codes: H04N 19/179 the unit being a scene or a shot
• CPC Class Codes: H04N 19/503 involving temporal prediction adaptive coding with adaptive selection between spatial and temporal predictive coding H04N19/107; adaptive coding with adaptive selection among a plurality of temporal predictive coding modes H04N19/109
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/87 involving scene cut or scene change detection in combination with video compression
• CPC Class Codes: H04N 5/147 Scene change detection

Abstract

An apparatus and a method of compressing and expanding a video signal. The apparatus includes a variable length coder controller and a quantizer. The coder controller compares a bit amount generated in the use of an optimum symbol with that in the use of an existing symbol upon detection of a scene change and outputs a new symbol upon the presence of a gain in accordance with the compared result. A frame delay delays input image data in a unit frame, a subtracter for subtracting an output signal of the frame delay from the input image data, and a scene change detector for detecting a scene change of the input image data. A motion detector detects a motion in the input image data, a main codebook for forming variable sub-codebooks for a still picture, a codebook controller for selecting a specified portion of the main codebook and variable sub-codebooks, and a vector formation unit for forming the input image data into a vector. A distortion measurement unit obtains a codeword with a minimum distortion using the variable sub-codebooks if the input image data is the still picture and using the main codebook if the input image data is a motion picture, and a code assignment unit for assigning a proper code to an output signal from the distortion measurement unit.

Claims

1. An apparatus for compressing and expanding a video signal, comprising:

motion estimation means having an input terminal for estimating a motion in the video signal inputted through the input terminal and generating a notion vector signal;

motion compensation means for compensating for the motion in the video signal in response to the motion vector signal;

subtraction means for subtracting an output signal of said motion compensation means from the video signal to output a difference signal;

discrete cosine transform means for performing a discrete cosine transform with respect to the difference signal from said subtraction means to remove spatial redundancy;

quantization means for compressing output information from said discrete cosine transform means, the output information from said discrete cosine transform means being approximated to a finite number of values to reduce an amount of the output information from said discrete cosine transform means;

inverse quantization means for restoring an output signal from said quantization means to an original state;

inverse discrete cosine transform means for transforming an output signal from said inverse quantization means to a spatial signal;

adding means for adding an output signal from said inverse discrete cosine transform means to the output signal from said motion compensation means and feeding back a resultant signal to said motion compensation means;

first mapping means for outputting information corresponding to the output signal from said quantization means;

scene change detection means for detecting a scene change in the video signal inputted through the input terminal;

variable length coder control means for stopping a learning occurring after the scene change and performing a new learning for a predetermined time period and allowing an optimum symbol to correspond to a given codebook, the variable length coder control means comparing a bit amount generated using the optimum symbol with a bit amount generated using an existing symbol in response to a scene change detect signal from said scene change detection means and a statistical characteristic of said changed scene, the variable length coder control means outputting a new symbol when a gain is present in accordance with a result of comparison of the bit amounts;

second mapping means for outputting information corresponding to said new symbol from said variable length coder control means;

first multiplexing means for selectively outputting the output information from one of said first and second mapping means;

variable length coder means for reducing an amount of the output information from said first multiplexing means without losing a characteristic of the output information from said first multiplexing means;

first buffering means for preventing one of an underflow and an overflow of an output signal from said variable length coder means resulting from transmission at a fixed bit rate; and

formatting means for formatting an output signal from said first buffering means to a suitable transmission format.

2. An apparatus for compressing and expanding a video signal according to claim 1, wherein said variable length coder control means includes:

a timing signal generator for generating a control timing signal at a fixed period in response to the scene change detect signal from said scene change detection means;

a symbol generator for generating symbols corresponding to run and level signals from said quantization means;

a histogram generator for generating a histogram according to frequencies of the symbols from said symbol generator and producing an output signal;

a sorter for sorting the symbols from said symbol generator according to a statistical value obtained for a predetermined learning time period and the frequencies of the symbols in response to the output signal from said histogram generator;

a rate calculator for comparing a bit amount generated using the optimum symbol with that using the existing symbol and outputting the new symbol when the gain is present in accordance with the compared result;

an index sender for indexing the new codeword in response to output signals from said sorter and rate calculator and sending an indexed new codeword to said formatting means; and

a downloader for downloading the indexed new codeword from said index sender to said second mapping means.

3. An apparatus for compressing and expanding a video signal according to claim 1, further comprising:

deformatting means for deformatting an output signal from said formatting means to a suitable reception format;

second buffering means for adjusting a number of bits of an output signal from said deformatting means at a desired rate;

shifting means for shifting an output signal from said second buffering means;

programmable logic array means connected to an output of said shifting means, said programmable logic array means storing a predetermined codeword;

third mapping means for outputting information corresponding to the predetermined codeword in said programmable logic array means;

control means for determining, in response to the output signal from said deformatting means, whether to use one of the new codeword and an existing codeword and outputting a select control signal in accordance with a discriminated result;

fourth mapping means for outputting information corresponding to the new codeword from said programmable logic array means; and

second multiplexing means for selectively outputting the output information from one of said third and fourth mapping means in response to the select control signal from said control means.

4. An apparatus for compressing and expanding a video signal according to claim 1, wherein said quantization means includes:

a frame delay for delaying input image data in a unit frame;

a subtracter for subtracting an output signal of said frame delay from the input image data to generate a difference signal;

a scene change detector for comparing the difference signal from said subtracter with a scene change threshold value to detect a scene change in the input image data;

a motion detector for comparing the difference signal from said subtracter with a motion estimation threshold value to detect a motion in the input image data;

a main codebook for forming variable sub-codebooks for a still picture in response to output signals from said scene change detector and motion detector;

a codebook controller for performing a control operation to select a specified portion of said main codebook and variable sub-codebooks in response to the output signals from said scene change detector and motion detector;

a vector formation unit for converting the input image data to a vector;

a distortion measurement unit, responsive to the output signals from said scene change detector and motion detector and an output signal from said vector formation unit, for obtaining a codeword with a minimum distortion using said variable sub-codebooks when the input image data is the still picture and using said main codebook when the input image data is a motion picture; and

a code assignment unit for assigning a proper code to an output signal from said distortion measurement unit.

5. An apparatus for compressing and expanding a video signal according to claim 4, wherein said variable sub-codebooks have a size smaller than that of said main codebook.

6. An apparatus for variable length coding of a video signal depending on image characteristics using image compression coding means which are comprised of well-known devices and operating functions including discrete cosine transform, quantization, motion compensation, motion estimator, inverse discrete cosine transform, and inverse quantization, the apparatus comprising

scene change detection means for detecting a scene change in the video signal inputted through an input terminal;

variable length coder control means for stopping a learning occurring after the scene change and performing a new learning for a predetermined time period and allowing an optimum symbol to correspond to a given codebook, the variable length coder control means comparing a bit amount generated using the optimum symbol with a bit amount generated using an existing symbol in response to a scene change detect signal from said scene change detection means and a statistical characteristic of said changed scene, the variable length coder control means outputting a new symbol when a gain is present in accordance with a result of comparison of the bit amounts;

first mapping means for outputting information corresponding to an output signal from image compression coding means;

second mapping means for outputting information corresponding to said new symbol from said variable length control means;

multiplexing means for selectively outputting the output information from one of said first and second mapping means; and

variable length coder means for reducing an amount of the output information from said multiplexing means without losing a characteristic of the output information from said multiplexing means.

7. An apparatus for expanding a video signal in response to a received video compression signal, the apparatus comprising:

deformatting means for deformatting an output signal from said received video compression signal to a suitable reception format;

buffering means for adjusting a number of bits of an output signal from said deformatting means at a desired rate;

shifting means for shifting an output signal from said buffering means;

programmable logic array means connected to an output;

of said shifting means, said programmable logic array means storing a predetermined codeword;

first mapping means for outputting information corresponding to the predetermined codeword in said programmable logic array means;

control means, responding to the output signal from said deformatting means for determining, whether to use one of a new codeword or the predetermined codeword, the control means outputting a select control signal in accordance with a determination result;

second mapping means for outputting information corresponding to the new codeword; and

multiplexing means for selectively outputting the output information from one of said first and second mapping means in response to the select control signal from said control means.

8. An apparatus for compressing and expanding a video signal, comprising:

motion estimation means having an input terminal for estimating a motion in the video signal inputted through the input terminal and generating a motion vector signal;

motion compensation means for compensating for the motion in the video signal in response to the motion vector signal;

subtraction means for subtracting an output signal of said motion compensation means from the video signal to output a difference signal;

discrete cosine transform means for performing a discrete cosine transform with respect to the difference signal from said subtraction means to remove spatial redundancy;

vector quantization means for compressing output information from said discrete cosine transform means, the output information from said discrete cosine transform means being approximated to a finite number of values to reduce an amount of the output information from said discrete cosine transform means;

inverse vector quantization means for restoring an output signal from said quantization means to an original state;

inverse discrete cosine transform means for transforming an output signal from said inverse vector quantization means to a spatial signal;

adding means for adding an output signal from said inverse discrete cosine transform means to the output signal from said motion compensation means and feeding back a resultant signal to said motion compensation means;

first mapping means for outputting information corresponding to the output signal from said quantization means;

scene change detection means for detecting a scene change in the video signal inputted through the input terminal;

variable length coder control means for comparing a bit amount generated using the optimum symbol with a bit amount using an existing symbol in response to a scene change detect signal from said scene change detection means, the variable length coder control means outputting a new symbol when a gain is present in accordance with a result of comparison of the bit amounts;

second mapping means for outputting information corresponding to a new codeword from said variable length control means;

first multiplexing means for selectively outputting the output information from one of said first and second mapping means;

variable length coder means for reducing an amount of the output information from said first multiplexing means without losing a characteristic of the output information from said first multiplexing means;

first buffering means for preventing one of an underflow and an overflow of an output signal from said variable length coder means resulting from transmission at a fixed bit rate; and

formatting means for formatting an output signal from said first buffering means to a suitable transmission format.

9. The apparatus of claim 8 wherein the variable length coder control means stops a learning occurring after the scene change and performs a learning for a predetermined time period and allows an optimum symbol to correspond to a given codebook, the variable length coder control means comparing a bit amount generated using the optimum symbol with a bit amount, generated using an existing symbol in response to a scene change detect signal from said scene change detection means and a statistical characteristic of said changed scene, the variable length coder control means outputting a new symbol when a gain is present in accordance with a result of comparison of the bit amounts.