Utilizing gray code and pixel decomposition to improve PPM compression performance

In digital gray-scale images, nearby pixels tend to have similar intensities, which normally are encoded using the ordinary binary code. However, the correlation between bits in such binary representation does not reflect how close these intensity values to each other. In this research, we proposed to utilize Gray code, instead of the ordinary binary code, and to decompose the image into multiple sub-images (two, four and eight). Each decomposed sub-image will be compressed separately using Prediction by Partial Matching (PPM) scheme, which is an adaptive statistical context-based lossless scheme. The proposed scheme is tested on 8 natural scene images. Results show improvement in compression performance when Gray code is utilized and the image is decomposed, compared to utilizing the ordinary binary code under the same conditions.