Method and apparatus for rate control for compression of video frames
View Patent ↗A method and apparatus for controlling the rate of compression in video/image compression environments is provided. The method includes determining a scaling factor that results in a compressed file size as close to a target file size as possible. In determining the scaling factor, tolerance limits of the current file size as compared with a predetermined tolerance is made, and when the tolerance is not met, the current scaling factor is interpolated to generate a new scaling factor, and consideration of a ratio of file sizes is made to determine if a new downsampling factor is needed for the interpolated scaling factor.
1. A method for rate control for compression of video frames in a sequence, the method comprising the steps of:
receiving a target compressed file size and file size tolerance;
compressing a subset of video frames in the sequence to obtain an estimate of compressed file size for all video frames in the sequence, wherein said compressing further comprises selecting a quantization table, a current scaling factor and a current downsampling factor for estimating said compressed file size for all frames in the sequence;
determining if an overall compressed file size is within a tolerance limit of said target compressed file size;
iteratively performing a scaling factor determination until the overall compressed file size is within the tolerance limit of said target compressed file size further comprising:
determining whether the target file size combined with the file size tolerance is less than the compressed file size, and when the determined target file size combined with the file size tolerance is less than the compressed file size,
setting a high file size to a compressed file size output from said step of compressing;
setting a low scaling factor to the current scaling factor; and
finding a lower bound step to produce a low file size and corresponding high scaling factor; and
selecting a scaling factor that results in a compressed file size approximating said target compressed file size.
2. A Method for rate control for compression of video frames in a sequence, the method comprising the steps of:
receiving a target compressed file size and file size tolerance;
compressing a subset of video frames in the sequence to obtain an estimate of compressed file size for all video frames in the sequence, wherein said compressing further comprises selecting a quantization table, a current scaling factor and a current downsampling factor for estimating said compressed file size for all frames in the sequence;
determining if an overall compressed file size is within a tolerance limit of said target compressed file size;
iteratively performing a scaling factor determination until the overall compressed file size is within the tolerance limit of said target compressed file size further comprising:
determining whether the target file size combined with the file size tolerance is less than the compressed file size, and when the determined target file size combined with the file size tolerance is greater than the compressed file size,
setting a low file size to a compressed file size output from said step of compressing;
setting a high scaling factor to the current scaling factor; and
finding an upper bound step to produce a high file size and corresponding low scaling factor; and
selecting a scaling factor that results in a compressed file size approximating said target compressed file size.
3. An apparatus for rate control for compression of video frames, the apparatus comprising:
circuitry that specifies a target compressed file size and file size tolerance;
a compressor that compresses a subset of a total number of video frames to obtain an estimate of compressed file size for all frames in a sequence, wherein said compressor selects a quantization table, a current scaling factor and a current downsampling factor for use in said compressing;
circuitry that determines if an overall compressed files size is within a tolerance limit;
a processor configured to iteratively perform a scaling factor determination when the overall compressed file size is not within the tolerance limit, wherein said processor is further configured to:
determine whether the target file size combined with the file size tolerance is less than the compressed file size, and when the determined target file size combined with the file size tolerance is less than the compressed file size,
set a high file size to a compressed file size output from said step of compressing;
set a low scaling factor to the current scaling factor; and
find a lower bound step to produce a low file size and corresponding high scaling factor; and
a selector to generate a scaling factor that results in a compressed file size as close to a target file size as possible.
4. An apparatus for rate control for compression of video frames, the apparatus comprising:
circuitry that specifies a target compressed file size and file size tolerance;
a compressor that compresses a subset of a total number of video frames to obtain an estimate of compressed file size for all frames in a sequence, wherein said compressor selects a quantization table, a current scaling factor and a current downsampling factor for use in said compressing;
circuitry that determines if an overall compressed files size is within a tolerance limit;
a processor configured to iteratively performing a scaling factor determination when the overall compressed file size is not within the tolerance limit, wherein said processor is further configured to:
determine whether the target file size combined with the file size tolerance is less than the compressed file size, and when the determined target file size combined with the file size tolerance is greater than the compressed file size,
set a low file size to a compressed file size output from said step of compressing;
set a high scaling factor to the current scaling factor; and
find an upper bound step to produce a high file size and corresponding low scaling factor; and
a selector to generate a scaling factor that results in a compressed file size as close to a target file size as possible.