Efficient compression of monochrome still images on general purpose hardware

The developed techniques of coding monochrome still images are reviewed. The most effective coding, in terms of decorrelating the image data, is found in the transform domain but due to the very high computational cost it is considered inappropriate on general purpose hardware. In the spatial domain, two techniques, Micro-Adaptive Picture Sequencing (MAPS) and Block Truncation Coding (BTC), offer the best performance in terms of computational cost and achievable compression. It is concluded that MAPS has a better fidelity performance at a cost of greater complexity. The BTC algorithm is the most efficient, in terms of computational cost, and has the potential for increased performance.