基于位平面分解方法的运动模糊图像PSF参数辨识

在基于匀速直线运动的运动模糊图像的复原中,点扩散函数(PSF)的两个参数(运动模糊角度和运动模糊尺度)的估计是研究的重点.为了能够准确地估计出PSF的模糊角度,提出了在一种新的改进倒频谱域中采用位平面分解提取和Radon变换相结合的方法.并在已知运动模糊角度的基础上再对模糊图像采用差分自相关处理,得到运动模糊尺度.将上述整个算法用Matlab进行仿真实验.实验结果表明:该算法在获取运动模糊参数的效率和准确度方面较现有文献的同类算法有所提高,得到的复原图像也更加清晰.

PSF Parameter Identification of Motion Blur Images Based on Bit-plane Segment Decomposition

During the restoration of the motion blurred image based on the uniform linear motion,the focus of the research is to estimate the two parameters (the motion blurred angle and the motion blurred scale) of the point spread function(PSF).In order to estimate the motion blurred angle of the PSF,a method adopting bit-plane decomposition and the Radon transform in a new improving cepstrum domain was proposed.And then,on the basis of the known motion blurred angle,the motion blurred image was dealt with the method of the differential autocorrelation in order to obtain the motion blurred scale.The whole algorithm was simulated with Matlab experiments.The experimental results show that the method to obtain motion blur parameters in efficiency and accuracy is improved compared with the similar algorithms in the existing literatures,besides,the obtained image restoration results are also clearer.