Revised Adaptive Threshold IFS Image Compression Based on Generalized Creditability
改进的广义置信度自适应IFS图象压缩编码算法

In this paper, the MSE in IFS image compression is analyzed. The concept of generalized creditability is presented. Based on that, the algorithm of the adaptive threshold(AT) IFS image compression using quadrature partitioning structure is proposed. To improve the compression ratio of the AT algorithm while still keeping the visual performance of decoded image, the formula of the adaptive threshold is revised according to the relative complexity of each range block, which forms the revised adaptive threshold(RAT) algorithm. The methods proposed in this paper set the threshold of current range block according to its statistics character, that is, variability. Therefore, the encoding process is adaptive to the complexity of the input image. Experiments results of algorithms based on AT, RAT and fixed threshold are given in this paper as comparison. The results show that RAT algorithm can compress the input image adaptively and the compression efficiency is improved considerably. Also time consumption of each algorithm is discussed at the end of this paper.