Fast Attractor Image Encoding by Adaptive Codebook Clustering

In attractor image compression systems, the encoders are often computationally intensive. An attractor encoder involves comparisons between two sets of blocks, called range blocks and domain blocks. The most basic type of encoder compares these sets exhaustively in the sense that every possible pair of range and domain domain blocks is examined. Unless the complexity of this task can be reduced, attractor coding may be excluded from many applications. This problem is recognized, and previous publications have presented strategies to overcome it.