摘要
根据运动造成图像模糊的特点,本文阐述了匀速直线运动模糊图像的退化模型,介绍了维纳滤波复原方法,提出了一种改进的运动模糊图像复原算法。该算法首先加强Radon变换算法的抗噪性能,利用自相关运算估计点扩展函数的模糊尺度参数,采用改进的K值自动估计算法较准确地估计出K值,再用维纳滤波复原图像。结果表明,这种改进的综合算法能够较为精确地估算出运动模糊图像的模糊参量,取得了较好的复原效果。
According to the characteristics of images blurring caused by motion, the degradation model of blurred images caused by uniform linear motion is analyzed in this paper. We introduce a method of the Wiener filtering recovery. The algorithm can first enhance the anti-noise performance of the Radon transform and use auto-correlation computing to estimate the point spread function of fuzzy scale, and then use the improved method of the K＇s automatic estimation algorithm to estimate a more accurate K. Then we use the Wiener filter to recover the blurred images. Experiments show that the integrated algorithm can estimate the fuzzy parameters of the motion-blurred images more accurately and obtain better recovery results.
出处
《计算机工程与科学》
CSCD
北大核心
2009年第10期49-52,共4页
Computer Engineering & Science
基金
湖南省自然科学基金资助项目(09JJ3115)
关键词
运动模糊图像
图像复原
Radon变换改进算法
自相关
motion-blurred image
image restoration
improved Radon transform algorithm
autocorrelation
