Visual Measuring Image Compression Based on Listless SPIHT Encoding and the Best Wavelet Packet Tree Structure

This paper presents a novel method of visual measuring image compression based on wavelet packet decomposition. Firstly, the high frequency sub-band decomposed by the wavelet transform is further decomposed into multiple directional sub-bands, so as to present the edge of the image more sparsely and accurately. Secondly select the best wavelet packet tree structure through comparing cost function among sub-bands, which is in compliance with the feature of visual measuring image. Thirdly in order to protect almost all detailed information, high-frequency was be emphasized according to the standard deviation of every sub-band. Last but not least, combined listless SPIHT encoding not only save memory space but also could be implemented by hardware conveniently. The experimental result shows that the novel compression method of visual measuring image has better capability of compression and maintaining accuracy.