A DCT Approximation with Low Complexity for Image Compression

This paper introduces an orthogonal approximation for the 8 point Discrete Cosine Transform (DCT). The proposed transformation matrix contains only ones and zeros. Bit shift operations and multiplication operations are absent. The approximate transform of DCT is obtained to meet the low complexity requirements. The implied transformation and approximation are orthogonal and are based on polar decomposition methods. The low complexity introduced in DCT reduces power consumption. The proposed image compression algorithm is comprehended using Matlab code. The proposed algorithm is less complex than Signed Discrete Cosine Transform (SDCT) and it requires two additional operations when compared to Bouguezel-Ahmed-Swamy (BAS) series of algorithms.