基于二维小波变换的图像压缩新算法

A New Image Compression Algorithm Based on Two-dimensional Wavelet Transform

为了实现快速、高质量的图像传输,用尽可能少的小波系数高质量地表征图像,提出了一种基于二维小波变换的图像压缩算法。首先将图像进行二维小波分解,得到各子带的一系列小波系数;其次保留图像低频子带的所有小波系数;然后根据归一化能量序列熵准则提取其他各层子带重要系数,即在某一门限下剩余系数熵值与剩余系数最大熵值比值较大(0.9左右)时,可以忽略这些小波系数,从而实现数据压缩;最后将各子带重要系数进行量化、编码。接收端根据接收到的数字信号重构图像信号。实验中对图像信号进行了压缩实验,验证了所提算法的正确性和有效性。实验结果表明本算法简单可行、搜索量小,为图像压缩找到了一种新的有效方法,具有一定的实际应用价值。

In order to realize high-speed image transmission with high-quality and use minimum wavelet coefficients to display the image,an image compression algorithm is proposed based on two-dimensional wavelet transformation.Firstly,the image is decomposed with two-dimensional wavelet to obtain a series of wavelet coefficients for each subband.Secondly,all wavelet coefficients of low-frequency subband are retained.Thirdly,important coefficients are extracted according to entropy criterion of normalized energy sequence.When the ratio of remainder coefficients＇entropy value and its maximum entropy value is relatively big（0.9 or so）,these remainder coefficients can be neglected and data compression is realized.Finally,important coefficients of each subband are quantized and encoded.In the receiving terminal the image is reconstructed according to received digital signal.Experiments with compressed image signals verify the correctness and effectiveness of this algorithm.The results have shown that this algorithm is simple and feasible,and its search workload is small.It provides a new effective method for image compression and has practical application value.