ST-ACO: Image Compression Using a New Adaptive Self-Organizing Tree Approach

This investigation presents an adaptive dynamic path selection algorithm (DPTSVQ) based on a self-organizing tree (S-TREE) using the threshold validity, called ST-ACO. ST-ACO employs an ant colony optimization framework (ACO) to adapt the nodes' threshold value incrementally. Furthermore, a fixed number of paths might impede self-organization, and result in searching on trap nodes. Experimental results indicate that the proposed algorithm not only generates better-quality decoded images than the S-TREE DoublePath algorithm, but also produces fewer candidate nodes than the MultiPath algorithm. Thus, the ST-ACO contributes hierarchical clusters, reducing the binary tree search bias by dynamic path searching and the adaptive threshold value in each node.