Enhanced approximate discrete cosine transforms for image compression and multimedia applications

Energy compaction property of the Discrete Cosine Transform (DCT) leads its usage in image and video compression applications. Nowadays power consumption is the major problem in modern multimedia applications. Hence circuit complexity and area reduction are the optimal ways for the power consumption reduction. In this paper the proposed DCTs represented as approximate DCT-Case I, II and III are constructed with basis elements C-i = {0, +/- 1, +/- 2}. Proposed Approximate DCT Case I and II have been derived from 16 x 16 point DCT rounding off operation and odd row sign replacement. Approximate DCT Case III has been derived based on (Potluri et al. in Meas Sci Technol 23(11):1-15, 2012). Peak signal to noise ratio (PSNR), total error energy and computational complexity wise approximate DCT Case III improved when compared to (Potluri et al. in Meas Sci Technol 23(11):1-15, 2012). Proposed approximate DCT Case I, II and III 8 x 8 matrixes satisfy the orthogonal property through a quantization process.