图像复原算法的频谱恢复特性分析研究

基于高斯函数假设的频谱分析方法,对逆滤波法、约束最小平方滤波法和最大似然法等三种图像复原算法的频谱恢复特性进行了实验和分析。假设光学传递函数H和图像频谱函数G为高斯函数,利用方差和方差比作为频谱宽度的分析评价指标,这是一种新的基于图像复原方法的频谱恢复特性分析法。在实验分析中,对H和G曲线设定两组方差,分为无噪声和有噪声两种情况,计算出各种算法的频谱曲线■及其方差和方差比。对由曲线图和数据给出的定性和定量结论进行了分析。

Analysis for spectrum restoration property of image restoration algorithms

Based on Gaussian function supposition,the image spectrum restoration property of the three image restoration algorithms is analyzed,such as the inverse filter,the constrained least squares restoration (CLS) and maximum likelihood estimation (PML). It is supposed that the optical transfer function H and the image spectrum function G are Gaussian functions,the variance and the variance ratio used as indexes of the spectrum width are analyzed and evaluated,it is a new method of analysis for image spectrum restoration property of the image restoration algorithm. The both curves H and G are enacted in two groups of different variances,and the image spectrum curve F is restored by the algorithms,its variance and the variance ratio are calculated in both cases (no-noise and with-noise).