An efficient DCT-based image compression system based on transparent composite model

Abstract

Recently, a new statistical model called the Laplacian transparent composite model (LPTCM) was developed for DCT coefficients, which could identify outlier coefficients in addition to providing superior modeling accuracy. In this paper, we aim at exploring its applications to image compression. To this end, we propose an efficient DCT-based non-predictive image compression system, where quantization and entropy coding are completely re-designed based on the LPTCM. Experimental results show that our proposed system (1) achieves the best rate distortion (RD) coding performance among all non-predictive image compression systems proposed in the literature and compared in the paper, and (2) provides, in comparison with HEVC intra coding, the state-of-the-art predictive image coding, comparable or even better RD coding performance in the highrate region or for complicated images, but with only less than 5% of the encoding complexity of the latter.