Runlength-Based Processing Methods for Low Bit-depth Images

Many low bit-depth images are compactly expressed in terms of runlength-based representations. This paper discusses methods for processing such images directly in runlength-represented form and how this can improve processing times for image processing operations. The primary focus in the analysis is on image compression operations, but an explanation of the overall scheme for runlength-based processing is also presented. Tradeoffs between processing times and compression ratio for runlength-based image compression algorithms are discussed. Sketches of runlength-based compression algorithms for some widely used image standards are given, and the algorithms are compared with raster-based counterparts. Important design features needed in implementations of runlength-based image processing are discussed. The main application areas of runlength-based image processing are also explained.