【摘要】 本文介绍了最小熵方法并引入应用于图像恢复技术中。文中考虑了变量可分离退化核。在退化点扩散函数(PSF)未知时,以差分图像的熵最小为准则直接估计恢复逆滤波器系数。如果退化图像上迭有加性噪声,差分模式则应作适当修改以减小噪声对参数估计的影响,然后以维纳滤波恢复图像。更多还原

【Abstract】 A minimum entropy deconvolution algorithm for restoration of sharp edge and step edge images is presented. The unknown PSF is estimated based on the criterion of minimum entropy of the difference image data. For noisy images, the difference model should be amended to decrease the estimation sensation due to noise. A Wiener’s filter is then applied to restore the image when the function of the inverse filter is available.更多还原