APPROXIMATION OF CALCULATIONS FOR FORWARD DISCRETE COSINE TRANSFORM

This paper presents new schemes to reduce the computation of the discr ete cosine transform (DCT) with negligible peak-signal-to-noise ratio (PSNR) degradation. The methods can be used in the software implementa tion of current video standard encoders, for example, H.26x and MPEG. We investigated the relationship between the quantization parameters a nd the position of the last nonzero DCT coefficient after quantization , That information is used to adaptively make the decision of calculat ing all 8 x 8 DCT coefficients or only part of the coefficients. To fu rther reduce the computation, instead of using the exact DCT coefficie nts, we propose a method to approximate the DCT coefficients which lea ds to significant computation savings. The results show that for pract ical situations, significant computation reductions can be achieved wh ile causing negligible PSNR degradation. The proposed method also resu lts in computation savings in the quantization calculations.