Adaptive Color Image Compression Using ADJPEG and ISUQ of Hierarchical Decomposition Scheme

This paper introduced a lossy color compression system of transform coding (TC) based of discrete wavelet transform (DWT), discrete cosine transform (DCT) and quantization schemes to achieve high compression ratio (CR) with preserving quality. The proposed compression system comprises the following steps, firstly, separating the image into source/non source color bands, then quantizing the source band uniformity, followed by decomposing an image by a three-level DWT and applying huffman coding to the approximation sub band, and compressed the details sub bands of each level by iterative scalar uniform quantization (ISUQ), while the non-source bands apply adaptive developed JPEG (ADJPEG) is utilized with the minimize matrix size algorithm (MMSA) and two integer keys base to reduce AC coefficients effectively. For testing the performance of the suggested compression system, three standard images of size (256×256) pixels adopted. The suggested technique showed superior performance in terms of reconstructed (decoded) image quality and CR, where the CR is between 28-32 with a peak signal-to-noise ratio (PSNR) value between 39-42 dB and the CR of JPEG is between 13-16 with a PSNR value between 33-37 dB.