Lossless/near lossless compression using bilateral filter and symmetrical structure based bias cancellation

Recently, a few symmetrical predictor structure (SPS) based lossless / near lossless compression algorithms have been proposed, which can efficiently exploit the information from the neighboring pixels. Prediction stage of existing SPS algorithms uses least squares optimization, which is computationally expensive and only causal pixels are used for the bias cancellation stage. In this paper, we propose an efficient and computationally simple lossless / near-lossless compression algorithm using the bilateral filter. Moreover we propose to use both causal and non-causal pixels for bias cancellation. From extensive experiments, it is observed that the proposed algorithm has the capability of provide better prediction and compression performance.