Methods and apparatus for reducing blockiness in decoded video

• US 5,654,759 A
• Filed: 02/15/1995
• Issued: 08/05/1997
• Est. Priority Date: 02/15/1995
• Status: Expired due to Term

• Current Assignee: Hitachi America Limited (Hitachi, Ltd.)
• Original Assignee: Hitachi America Limited (Hitachi, Ltd.)
• Inventors: Boyce, Jill MacDonald, Augenbraun, Joseph E.
• Primary Examiner(s): Chin, Tommy P.
• Assistant Examiner(s): RAO, ANAND SHASHIKANT
• Application Number:
  US08/388,769
• Time in Patent Office: 902 Days
• Field of Search: 348/405, 348/406, 348/409, 348/411, 348/416, 348/419, 348/420, 348/699, 348/404, 348/384
• US Class Current: 375/240.03
• CPC Class Codes: H04N 19/86 involving reduction of coding artifacts, e.g. of blockiness
• CPC Class Codes: H04N 5/21 Circuitry for suppressing or minimising disturbance, e.g. moiré or halo

Abstract

Methods and apparatus directed to reducing blockiness resulting from encoding digital video data using, e.g., DCTs are disclosed. An encoder which encodes video signals into a digital video bitstream including information identifying blocky video and instructions for processing the blocky video is also disclosed. A plurality of methods of identifying blocky video in an encoded bitstream by analyzing the encoded bitstream are also disclosed. In addition, various methods of filtering both non-blocky and blocky video, without adversely affecting the non-blocky video, are described.

Claims

1. A method of operating a video decoder, comprising the steps of:

receiving encoded digital video data, the encoded video data representing blocky and non-blocky predictively coded video frames and including motion vectors required for the decoding of the encoded video data;

monitoring the received encoded digital bitstream to identify the motion vectors required for the decoding of the encoded video data;

identifying video data corresponding to blocky video frames by

detecting motion vectors, indicative of blocky video frames, that are larger than a preselected minimum size; and

decoding the received encoded video data using the motion vectors included in the received encoded video data, the decoding step including the step of performing a blockiness reduction operation on video frames which are decoded using detected motion vectors that are larger than the preselected minimum size.

2. A method of operating a video decoder, comprising the steps of:

receiving an encoded digital bitstream including encoded digital video data corresponding to blocky and non-blocky video images, the encoded digital video data corresponding to blocky and B non-blocky predictively coded video frames, each predictively coded video frame including intra-coded blocks and predictively coded blocks required for the decoding of the encoded video data;

monitoring the received encoded digital video data to identify the intra-coded blocks; and

identifying video data corresponding to blocky video images by detecting video data corresponding to predictively coded video frames that include more than a preselected minimum percentage of intra-coded blocks.

5. The method of claim 2, further comprising the steps of:

decoding the encoded digital video data to generate decoded video data; and

processing the decoded video data generated by decoding the encoded digital video data identified as corresponding to blocky video images by performing a Gaussian low pass filter operation thereon.

6. The method of claim 2, further comprising the steps of:

decoding the encoded digital video data to generate decoded video data including blocks having block boundaries; and

processing the decoded video data generated by decoding the encoded digital video data identified as corresponding to blocky video images by performing a low pass filter operation across the block boundaries.

3. A method of operating a video decoder comprising the steps of:

receiving encoded digital video data corresponding to blocky and non-blocky video images, the encoded video data including blocks of video data required for the decoding of the encoded video data, each block corresponding to an array of pixels;

monitoring the received encoded digital video data to identify the blocks of video data required for the decoding of the encoded video data;

identifying video data corresponding to blocky video images by using information included in the blocks of video data to calculate a blockiness cost function using pixel differences at block boundaries; and

decoding the received encoded video data, the decoding step including the step of performing a blockiness reduction operation on the video data identified as corresponding to blocky video images.

4. The method of claim 3,

wherein the blockiness cost function is calculated for at least three adjacent blocks, represented by the letters A, B, and C, block A, block B and Block C each representing an array of 8×

8 pixels, and

wherein the cost function is expressed using the following equation;

##EQU3## where A(0,Y) refers to the (0,Y) pixels in the array of 8×

8 pixels which comprise block A;

where B(7,Y) refers to the (7,Y) pixel in the array of 8×

8 pixels which comprise block B;

where A(X,0) refers to the (X,0) pixel in the array of 8×

8 pixels which comprise block A; and

where C (X,7) refers to the (X,7) pixel in the array of 8×

8 pixels which comprise block C.

7. A method of operating a video decoder comprising the steps of:

receiving an encoded digital bitstream including encoded digital video data corresponding to blocky and non-blocky video images, the encoded video data including information required for the decoding of the encoded video data;

monitoring the received encoded digital bitstream to identify the information required for the decoding of the encoded video data;

using the information required for the decoding of the encoded video data to identify video data corresponding to blocky video images;

decoding the encoded digital video data to generate decoded video data including predictively coded blocks having blocking artifact boundaries located within the predictively coded blocks; and

performing a blockiness reduction operation on at least a portion of the decoded video data by performing a low pass filter operation across the blocking artifact boundaries located within the predictively coded blocks.

8. A method of operating a video decoder comprising the steps of:

receiving an encoded digital bitstream including encoded digital video data corresponding to blocky and non-blocky video images, the encoded digital video data including blocks of video data, each block corresponding to an array of pixels;

monitoring the received encoded digital bitstream to identify encoded digital video data corresponding to blocky video images, the monitoring step including the step of;

calculating a blockiness cost function using pixel differences at block boundaries;

decoding the encoded digital video data to generate decoded video data, the decoding step including

the step of performing an IDCT operation on a block represented by DCT coefficients; and

processing the decoded video data generated by decoding the encoded digital video data identified as corresponding to blocky video images by performing a blockiness reduction operation thereon,

the blockiness reduction operation including the steps of;

altering at least one DCT coefficient of the block; and

calculating the IDCT for the block.

9. The method of claim 8, wherein the at least one DCT coefficient of the block is altered to minimize the blockiness cost function.

10. A method of processing video data representing predictively coded images and bi-directionally coded images that were generated by performing motion compensation on a video signal, the method comprising the steps of:

identifying video data corresponding to blocky video images by detecting motion vectors having a value exceeding a preselected minimum motion vector value indicative of blocky video images; and

performing a blockiness reduction operation on the identified blocky video data when the preselected minimum motion vector value is exceeded.

11. A method of processing digital video data including predictively coded video frames represented by intra-coded and predictively coded blocks, the method comprising the steps of:

monitoring the number of intra-coded blocks in the predictively coded video frames;

identifying blocky predictively coded video frames by;

i. monitoring the number of intra-coded blocks in the predictively coded video frames; and

ii. detecting which ones of the predictively coded video frames include more than a preselected minimum number of intra-coded blocks, the detected predictively coded video frames; and

performing a blockiness reduction operation on the detected predictively coded video frames.

12. A method of detecting and processing blocky video images in a series of encoded video images, the method comprising the steps of:

examining video images in the series of encoded video images to identify fade or scene changes indicative of blocky video images; and

performing a blockiness reduction on the video images in which fade or scene changes are identified.

13. A method of performing a blockiness reduction operation on blocky decoded video data including predictively coded blocks having blocking artifact boundaries located within the predictively coded blocks, the method comprising the step of:

performing a filtering operation across the blocking artifact boundaries located within the predictively coded blocks of the blocky decoded video data.

14. The method of claim 13, wherein the filtering operation is a low pass filtering operation.