摘要： 为提高含噪运动图像的复原质量，提出一种基于互相关原理的运动模糊长度估计算法。根据图像的Radon变换得到运动模糊角度，提取模糊中心各模糊路径上的像素灰度值，通过互相关计算确定模糊长度，采用Lucy-Richardson迭代算法对观测图像进行复原。实验结果表明，该算法的抗噪性较好，迭代收敛速度较快。

关键词: 图像复原, 互相关, Radon变换, 点扩散函数, 模糊长度, 运动模糊

Abstract: In order to improve the quality of image restoration for motion blur images with noise signal, a new motion blur length estimation algorithm based on the principle of the correlation is proposed in the paper. To find motion blur parameters, Radon transform method is used to estimate motion blur angle. It extracts gray value of pixels which is around blur center, calculates correlation for getting motion blur length, and uses Lucy-Richardson iterative algorithm to restore the degraded image. Experimental results show that the algorithm has better antinoise and faster iterative convergence speed.

Key words: image restoration, cross-correlation, Radon transform, Point Spread Function(PSF), blur length, motion blur

中图分类号: 

• TN911.73