基于小波变换的图像压缩

Image Compression Based on Wavelet Transform

摘要: 近年来，随着数字通信技术的发展和网络、多媒体技术的兴起，图像编码与压缩作为数据压缩的一个分支收到越来越多的关注。小波压缩技术是一种有损的图像压缩技术，利用Mallat分解算法或者小波包方法对图像进行压缩与重构可以在保留图像重要信息的条件下，尽量减少图像的存储空间，便于储存和传输。本文通过运用Haar小波和db4小波，两种小波对图像进行两层分解和低频系数重构，发现对同一张图像，同样层数的压缩Haar小波和db4小波对图像压缩的效果(压缩比)相差不大。而“sym4”小波包的压缩效果要比小波压缩效果更好。

Abstract: In recent years, with the development of digital communication technology and the rise of network and multimedia technology, image coding and compression, as a branch of data compression, have received more and more attention. Wavelet compression technology is a lossy image compression technology. Using Mallat decomposition algorithm or wavelet packet method to compress and reconstruct the image can minimize the storage space of the image and facilitate storage and transmission under the condition of retaining the important information of the image. In this paper, Haar wavelet and db4 wavelet are used to decompose the image into two layers and reconstruct the low-frequency coefficients. It is found that for the same image, the compression effect (compression ratio) of Haar wavelet and db4 wavelet with the same number of layers is almost the same. The compression effect of “sym4” wavelet packet is better than that of wavelet compression.