JPEG-Like Color Image Compression Based on All Phase Biorthogonal Transform and Improved Quantization Table

摘要

With the development of multimedia technologies, color image compression requires higher performance. To address this need and fit the human visual system (HVS), this paper adopts the improved luminance quantization table obtained by duality nonlinear programming and the uniform chrominance quantization table, which are applied to JPEG-like color image compression algorithm based on all phase biorthogonal transform (APBT). Besides, we also apply proposed compression algorithm to deal with color filter array (CFA) image data. All experimental results show that our algorithm ensures the slightest increase of computational complexity and improves the quality of the color image. The blocking artifacts also have been partly removed, especially at low bit rates. In general, the proposed algorithm can produce optimal rate-distortion and better subjective image quality compared with some existing methods.