Zero-Tree Wavelet Algorithm Joint with Huffman Encoding for Image Compression

Embedded Zero-tree Wavelet EZW is an effective image encoding algorithm. This paper emphasizes on the principles of EZW improved algorithm and the realization process for algorithm that includes zero-tree structure, wavelet coefficient scanning mode, improving embedding EZQ algorithm flow. Finally, Huffman coding was jointed to encoding. By carefully analyzing EZW algorithm, we set the edge threshold as a significant coefficient and querying it with maximum value to determine whether it's the zero-tree root or isolated zero. If the maximum is greater than the threshold, then it will be an isolated zero. The improved algorithm will replace the arithmetic coding method with Huffman coding making it more simpler. Finally, we simulated the improved algorithm in Matlab to validate the result. Our result shows that in comparison with independent EZW algorithm, the improved algorithm not only increases the compression ratio and encoding efficient, but also improves the peak signal to noise ratio of images and make the vision more clear. Hence proved that the improved algorithm is more feasible and effective.