Lossless image coding using a switching predictor with run-length encodings

Abstract

In this paper, we propose a switching adaptive predictor (FSWAP) with run-length encodings for lossless image coding. The proposed FSWAP system has two operation modes, run mode and regular mode. If the members in the texture context of the coding pixel have identical grey values, the run mode will be used; otherwise regular mode is used. The run mode using run-length coding with an arithmetic coder is very useful for image with flat regions. The regular mode borrows the switching predictor structure in SWAP [1] with some modifications. The SWAP coder, updating network weights on the fly and making automatic fuzzy context modeling, has been shown to provide very good results [1]. To make the proposed system more feasible for applications of limited resources, the dimension of ANP in SWAP [1] is simplified and the automatic fuzzy context modeling is replaced by a fixed context clustering. Experiments show that the simplified context clustering is very useful in modeling error for prediction refinement. Furthermore, the execution time of FSWAP can be accelerated with minor degradation in the bit rates associated with the modifications. Comparisons of the proposed system to existing state-of-the-art predictive coders will be given to demonstrate its coding efficiency.