基于改进R-L算法的运动模糊图像复原方法研究

Research of Motion Blur Image Restoration Method Based on Improved R-L Algorithm

通过对运动模糊图像的复原过程进行分析,提出了一种基于二次变换改进的路西理查德森(Richardson-Lucy,R-L)算法图像复原方法,利用二次变换法精确估计出点扩散函数的系数,结合R-L算法完成退化图像的复原过程。通过仿真实验,量化对比了上述方法与传统R-L算法的复原效果,所述方法复原图像的灰度平均梯度以及峰值信噪比均提高约30%,实验结果表明,有效提升了运动模糊图像的复原精度,且针对含噪图像具有更高的鲁棒性。

Based on the image point spread function(PSF)estimation and restoration algorithm,by analyzing the restoration process of motion blurred images,a method based on two-time transform parameter estimation and Richardson-Lucy(R-L)algorithm for image restoration were presented.Through simulation experiments,the advantages and disadvantages of the above method and the traditional image restoration methods were compared and analyzed,both the gray scale mean gradient and the Peak signal-to-noise ratio were increased by about 30%.The experimental results shown that the method can effectively improve the restoration accuracy of degraded images,and has higher robustness for noisy images.