[Paper] Lossless Color Image Coding Based on Probability Model Optimization Utilizing Example Search and Adaptive Prediction

We previously proposed a novel lossless coding method that utilizes example search and adaptive prediction within a framework of probability model optimization for gray-scale images. In this paper, we extend the method for RGB 4:4:4 formatted color images. In the proposed method, multiple examples are collected from the causal area in not only the same color signal to be encoded but also other color signals as far as they have already been encoded. Moreover, multiple affine predictors trained on a pel-by-pel basis are also utilized to exploit intra- and inter-color correlations. The probability distribution of the color signal at each pel is dynamically modeled by using both examples and predictors. Then a few parameters used in the probability model are numerically optimized for efficient entropy coding. The experimental results show that the proposed method achieves better coding performance than other state-of-the-art lossless coding methods.