针对图像压缩中小波基的选取自由度高的问题，讨论了小波基的选择依据，分析了正交性、对称性、正则性、支撑性和消失矩等性能指标对图像编码的影响。研究了利用均方误差和峰值信噪比对图像品质的评价方法。利用MATLAB仿真，采用SPIHT压缩算法，使用4种小波基对测试图像进行压缩，对比其均方误差、最大误差、峰值信噪比和压缩比，并对实验结果进行了分析，发现具有对称性和紧支撑性，正则性较好的双正交小波综合性能最好，为小波基的选择提供了参考标准。%Wavelet basis selection was discussed due to its high degree of freedom in image compression. The influence of orthogonality,symmetry,regularity,support and vanishing moments in image coding was analyzed. The method of image quality assessment using mean square error and peak signal to noise ratio was illustrated. Test images were compressed using four different wavelet basis through MATLAB,and mean square error,max error,peak signal to noise ratio and compress ratio were compared respectively. The conclusion is that wavelet basis which are symmetric,tight supported,good regularity and biorthogonal have the best comprehensive performance,which provides reference standard for the choice of wavelet basis.
展开▼
