摘要
研究了运动模糊图像方向角的估计方法,通过数学原理分析,以及编写matlab程序进行数值实验,对Hough变换、Radon变换和Cepstral变换三种方法进行了对比研究,综合比较了三种方法的估计精度,指出了每种方法的适用范围。结果表明,Hough变换依赖于特征点的选取,选取不好则估计失效;Radon变换最为实用,噪声干扰对其估计精度影响不大,而且运动距离大时估计更准确;Cepstral变换在噪声水平低的时候估计较好,但噪声较大时估计精度差。
In order to identify the motion direction for motion blurred images, three common methods including Hough transformation, Radon transformation and Cepstral transformation are analyzed in this paper. Based on theoretical a- nalysis and large number of numerical experiments through matlab programs, three methods are comprehensively eval- uated. Finally, the scope of application for each method is reached. The results show that, Hough transform is depend- ent on the selection of feature points;Radon transform is the most practical, the noise has little effect on estimation accuracy, and the estimation is better for the longer motion distance. The estimation by using Cepstral transform is ac- curate when the noise level is low ,and the greater the noise is ,the worse the estimation accuracy will be.
出处
《太原科技大学学报》
2012年第6期426-430,共5页
Journal of Taiyuan University of Science and Technology
基金
北京工业职业技术学院科研基金(2012bgzyky24)
