HVS-Optimized Vector Quantilizer for Remote Sensing Texture Compression

As remote sensing texture has properties like strong randomness and weak local correlation, it is hard to design a good vector quantizer for compression. A novel self-adaptive HVS-optimized quantizer is presented. The method defines a similarity measurement function based on human visual system (HVS) model. Threshold that judge the similarity between blocks is computed based on the property of image. Thus, the compression method may deal with different resolution images automatically. In addition, a self-adaptive threshold adjustment during the compression is designed to improve the reconstruct quality for remote sensing textures with large regional differences. Experiment on various resolution images indicates that the new quantizer can achieve satisfied compression rate and reconstruct quality at the same time. Furthermore, the compression and decompression process is speed up with the usage of GPU, on account of their parallelism.