控制工程 ›› 2019, Vol. 26 ›› Issue (12): 2159-2163.

• •    下一篇

基于Richardson-Lucy算法的离焦泡沫图像复原

  

  • 出版日期:2019-12-20 发布日期:2023-11-29

Richardson-Lucy Algorithm Based Defocused Bubble Images Restoring

  • Online:2019-12-20 Published:2023-11-29

摘要: 以基于机器视觉的选矿过程为背景,针对摄像机采集到的模糊泡沫图像,基于图像先验和Richardson-Lucy算法对模糊图像进行复原。首先基于图像先验在多尺度环境下估计图像的模糊核;然后为了抑制饱和区域对图像复原的影响,对Richardson-Lucy算法进行改进;最后利用估计的模糊核和改进的Richardson-Lucy算法对模糊图像进行图像复原。实验结果表明,和Richardson-Lucy算法及其他的算法相比,改进算法复原的图像具有更好的视觉效果和客观评价参数。

关键词: Richardson-Lucy算法, 梯度先验, 模糊核估计, 图像盲复原

Abstract: As for flotation process based on machine vision, blurred images are restored based on image priors and Richardson-Lucy algorithm, considering blurred bubble images acquired by industrial camera. Firstly, blur kernel is estimated based on image priors in multi-scale. Then the Richardson-Lucy algorithm is improved to suppress the influences of saturated regions. Finally, the blurred image is restored utilizing estimated blur kernel and improved Richardson-Lucy algorithm. Experimental results demonstrate that compared with the Richardson-Lucy algorithm and other algorithms, the improved algorithm can restore image with better visual results and objective evaluation.

Key words: Richardson-Lucy algorithm, gradient prior, blur kernel estimation, blind image restoration