Fixed rate intraframe compression and decompression of video based on visual quality

Patent metadata

Field	Value
Publication number	US-12506887-B2
Application number	US-202418940335-A
Country	US
Kind code	B2
Filing date	Nov 7, 2024
Priority date	Apr 12, 2024
Publication date	Dec 23, 2025
Grant date	Dec 23, 2025

Abstract

DCT-based methods for fixed bit rate, intraframe compression of video are based on the visual quality of the resulting decompressed imagery. In general, this involves assigning an initial bit budget per macroblock of a given video frame, resulting in degrees of compression, i.e., quantization scale factors, that vary among the macroblocks according to their complexity. The scale factors are then adjusted while maintaining the overall frame bit budget to reduce the visibility of artifacts in the decompressed frame. The adjustments may include increasing scale factors for simple macroblocks and reducing scale factors for complex blocks. As compared to traditional DCT-based compression schemes, the described methods may reduce the visibility of compression-related artifacts both in complex and in simple portions of the frame.

First claim

What is claimed is:

1 . A method for fixed bit rate, intraframe compression of video, including a sequence of images, comprising, for each image:
segmenting the image into n portions, wherein compression parameters are individually adjustable for each of the n portions; transforming each of the portions of the image to generate frequency domain coefficients for each of the portions; selecting a stored quantization matrix for the image based on a desired bit rate for the compressed video; determining an initial bit budget per transformed portion of the image that is equal to a bit budget for the image within the compressed video divided by n; for each transformed portion, determining an initial scale factor that is the smallest scale factor that, when applied to the frequency domain coefficients together with the quantization matrix, results in a bit size of the transformed portion that is equal to or less than the initial bit budget per transformed portion; and generating a bitstream representing the compressed image by:
for each transformed portion, scaling the portion's frequency domain coefficients by applying to the portion the initial scale factor determined for that portion together with the quantization matrix; and
arranging the scaled frequency domain coefficients in a predetermined sequence and applying variable length coding to the sequence of scaled frequency domain coefficients.

2 . The method of claim 1 , further comprising:
ranking the transformed portions in order of their initial scale factors and determining an average of the initial scale factors; proceeding in order of increasing initial scale factor, adjusting downward the initial scale factors of a plurality of transformed portions having initial scale factors greater than the average initial scale factor, wherein the amount of adjustment is a diminishing fraction of a difference between a given transformed portion's initial scale factor and the average initial scale factor; and terminating the adjustment of initial scale factors when adjusting an additional transformed portion would result in a sum of bit sizes of the scaled transformed portions of the image exceeding a bit budget for the image within the compressed video.

3 . The method of claim 2 , further comprising:
prior to adjusting downward the initial scale factors of the plurality of transformed portions having initial scale factors greater than the average initial scale factor, for each portion having an initial scale factor below a threshold scale factor, adjusting upward the initial scale factor of the portion, wherein the threshold scale factor is based on the average initial scale factor and a standard deviation of the initial scale factors.

4 . A computer program product comprising:
a non-transitory computer-readable medium with computer-readable instructions encoded thereon, wherein the computer-readable instructions, when processed by a processing device instruct the processing device to perform a method for fixed bit rate, intraframe compression of video, including a sequence of images, the method comprising, for each image:
segmenting the image into n portions, wherein compression parameters are individually adjustable for each of the n portions;
transforming each of the portions of the image to generate frequency domain coefficients for each of the portions;
selecting a stored quantization matrix for the image based on a desired bit rate for the compressed video;
determining an initial bit budget per transformed portion of the image that is equal to a bit budget for the image within the compressed video divided by n;
for each transformed portion, determining an initial scale factor that is the smallest scale factor that, when applied to the frequency domain coefficients together with the quantization matrix, results in a bit size of the transformed portion that is equal to or less than the initial bit budget per transformed portion; and
generating a bitstream representing the compressed image by:
for each transformed portion, scaling the portion's frequency domain coefficients by applying to the portion the initial scale factor determined for that portion together with the quantization matrix; and
arranging the scaled frequency domain coefficients in a predetermined sequence and applying variable length coding to the sequence of scaled frequency domain coefficients.

5 . A system comprising:
a memory for storing computer-readable instructions; and a processor connected to the memory, wherein the processor, when executing the computer-readable instructions, causes the system to perform a method for fixed bit rate, intraframe compression of video, including a sequence of images, the method comprising, for each image:
segmenting the image into n portions, wherein compression parameters are individually adjustable for each of the n portions;
transforming each of the portions of the image to generate frequency domain coefficients for each of the portions;
selecting a stored quantization matrix for the image based on a desired bit rate for the compressed video;
determining an initial bit budget per transformed portion of the image that is equal to a bit budget for the image within the compressed video divided by n;
for each transformed portion, determining an initial scale factor that is the smallest scale factor that, when applied to the frequency domain coefficients together with the quantization matrix, results in a bit size of the transformed portion that is equal to or less than the initial bit budget per transformed portion; and
generating a bitstream representing the compressed image by:
for each transformed portion, scaling the portion's frequency domain coefficients by applying to the portion the initial scale factor determined for that portion together with the quantization matrix; and
arranging the scaled frequency domain coefficients in a predetermined sequence and applying variable length coding to the sequence of scaled frequency domain coefficients.

Assignees

• Avid Tech Inc

Inventors

• Weber Markus

Classifications

• H04N19/625

  using discrete cosine transform [DCT] · CPC title

• H04N19/48

  using compressed domain processing techniques other than decoding, e.g. modification of transform coefficients, variable length coding [VLC] data or run-length data (motion estimation in a transform domain H04N19/547; processing of decoded motion vectors H04N19/513) · CPC title

• H04N19/184

  the unit being bits, e.g. of the compressed video stream · CPC title

• H04N19/136

  Incoming video signal characteristics or properties · CPC title

• H04N19/124 Primary

  Quantisation · CPC title