Improved Chroma Resampling Algorithm Based on Interpolation Method

Abstract

An improved algorithm based on bilinear interpolation was proposed. It could deal with the shortcoming of chroma component information loss in the original video image chroma resampling algorithm. With the bilinear interpolation method, the chroma component matrix in the YCbCr color space was scaled in both the horizontal and vertical directions. It could reduce information loss while meeting the corresponding standard data format. The new algorithm could retain more information from the source image’s color, which improved image recovery. In addition, the paper proposed a new format with a higher compression ratio on the basis of ensuring visual performance. Experimental results showed that, compared with the original chroma resampling algorithm, the processing result of the improved algorithm had a significant improvement in the performance of parameters such as composite peak signal noise ratio (CPSNR).