【摘要】 空间变化的运动模糊图像复原是当今图像处理领域的热点问题之一,而运动模糊区域检测更是其中的重点和难点.针对以往运动模糊区域检测算法的精确度较低的问题,提出基于检测分割的单帧运动模糊图像盲复原算法.结合多尺度感知和闭合型图分割,实现两步模糊区域检测.将检测到的模糊区域外延后进行复原,再将其去除外延部分并与清晰区域拼接,得到最终的复原结果.实验结果表明:本文算法不但显著提高了模糊区域检测精度,也有效提高了图像复原质量.同时,具有一定的普遍性,可以适用于自然场景下的多种空间变化的运动模糊图像.更多还原

【Abstract】 Nowadays restoration of spatial changed motion blurred image is one of the hot issues in the field of image processing,and motion blurred region detection is the key and difficult point.Due to the problem of low accuracy of previous motion blurred region detection algorithms,a blind restoration algorithm of the single motion blurred image based on blur detection and segmentation is proposed.The two-step blurred region detection is realized by combining multiscale perception and the closed-Form matting algorithm.Furthermore,the detected blurred region is expanded and restored.Then the extended part is removed and the restored blurred region is spliced with the clear region to obtain the final restored results.The experimental results show that the proposed algorithm not only significantly improves the accuracy of the blurred region detection,but also effectively improves the quality of image restoration.At the same time,it also has certain universality and can be applied to many kinds of spatial changed motion blurred images in natural scenes.更多还原