基于坐标变换的相向运动图像模糊复原方法

Restoration of Forward Motion Blurred Image Based on Polar Coordinates

针对递归计算自身产生的残差以噪声形式出现在复原后辐射状模糊图像上的问题,研究了一种通过坐标变换和空域点运算获得图像退化模型,进而对该模型求逆复原辐射状模糊图像的新方法.使用扇形网格进行了图像的重新采样,在此基础上,根据图像退化过程推导出了图像退化模拟矩阵;使用最小二乘法作为约束条件对重新采样的图像进行了复原.结果表明,该算法简单,复原图像清晰,可以用以进行图像辐射状模糊的复原.

Aiming at the problem of recursion computation errors present in the form of noise at deblurred image, the image degradation model being based on coordinate transform and dot operation in airspace, image restoration is here investigated on the counterreaction of image degradation. After sampling afresh with the sector gridding on the image, image degradation simulation matrix is deduced on the base of image degradation, and the resampled image with restriction conditions restored with least-square multiplication. The results showed that the algorithm was simple, providing a clearer restoration, so that the radially blurred image can be restored with algorithm given above.