A New Predictive Image Compression Scheme Using Histogram Analysis and Pattern Matching

In this paper, we introduce a new predictive image compression scheme that compresses an image by a set of parameters computed for individual blocks of different types. These parameters include the average and difference of the representative intensities of an image block, together with the index of a pattern associated with the block visual activity. The block representative gray values are computed through a histogram analysis of the block residuals and a pattern matching technique is employed to find the best match for the block bit-pattern from a pre-defined pattern book. To further reduce the bit rate, a predictive technique selectively predicts the parameters based on the corresponding values in the neighboring blocks. The simulation results confirm that the proposed technique can provide a high compression ratio with acceptable image quality of the compressed images.