A Lossless Image Compression Algorithm using Differential Subtraction Chain

This paper presents a novel algorithm for image compression aiming to reduce data packet size, resulting in effective bandwidth utilization during data transmissions. The algorithm referred to as Differential Subtraction Chain (DSC) consists of three steps. First, it separates an image file to three matrices of RGB. Second, it computes element-wise different values in each pixel between R and G matrices, and between G and B matrices. Third, the different values are binary encoded and transformed to consecutive vectors in order for data transmissions. In our MATLAB simulations, the performance measure is compression ratio which is calculated by [1-(compressed data size/original data size)] 100%. The compression ratios yielded by our DSC tested with three benchmarking images of city, Lenna and Mandrill are 44.02%, 42.02% and 39.86%, respectively. The result shows that DSC is much better than other existing algorithms in the experiment.