System and method for improving video data compression by varying quantization bits based on region within picture
Presented herein is a system and method for improving video data compression by varying quantization bits based on a region within picture. In one embodiment, there is presented a method for encoding video data. The method comprises dividing a picture into a plurality of blocks; compressing a particular one of the plurality of blocks with lossless compression; measuring how far the particular block is from a center of the picture; and compressing the particular one of the blocks with lossy compression, wherein information loss is based on how far the particular block is from the center of the picture.
1 . A method for encoding video data, said method comprising:
dividing a picture into a plurality of blocks;
compressing a particular one of the plurality of blocks with lossless compression;
measuring how far the particular block is from a center of the picture; and
compressing the particular one of the blocks with lossy compression, wherein information loss is based on how far the particular block is from the center of the picture.
2 . The method of claim 1 , wherein dividing the picture comprises:
dividing the picture into macroblocks.
3 . The method of claim 1 , wherein the lossless compression comprises representing the particular block with frequency coefficients.
4 . The method of claim 3 , wherein representing the particular with frequency coefficients further comprises:
predicting the particular block, from a prediction block;
calculating a prediction error between the block and the prediction block; and
transforming the prediction error to a frequency domain.
5 . The method of claim 3 , wherein the lossy compression comprises quantizing the frequency coefficients with bits, wherein the number bits is based on the how far the particular block is from the center of the picture.
6 . The method of claim 5 , wherein the number of bits quantizing the frequency coefficients is greater if the particular block is close to the center of the picture and fewer if the block is far from the center of the picture.
7 . A video encoder for encoding video data, said video encoder comprising:
a lossless compression engine for compressing a block of the picture with lossless compression;
a controller for measuring how far the particular block is from a center of the picture; and
a lossy compression engine for compressing the blocks with lossy compression, wherein information loss is based on how far the particular block is from the center of the picture.
8 . The video encoder of claim 7 , wherein the block comprises a macroblock.
9 . The video encoder of claim 7 , wherein the lossless compression engine comprises a transformation engine for representing the block with frequency coefficients.
10 . The video encoder of claim 9 , wherein the lossless compression engine comprises:
a motion estimator for predicting the particular block, from a prediction block;
a subtractor for calculating a prediction error between the block and the prediction block; and
wherein the transformation engine transforms the prediction error to a frequency domain.
11 . The video encoder of claim 9 , wherein the lossy compression engine comprises a quantizer for quantizing the frequency coefficients with bits, wherein the number of bits is based on the how far the particular block is from the center of the picture.
12 . The video encoder of claim 11 , wherein the number of bits quantizing the frequency coefficients is greater if the particular block is close to the center of the picture and fewer is the block is far from the center of the picture.