摘要
针对图像的局部模糊现象,提出基于四向差分的局部模糊图像检测算法。首先计算待检测图像的在0°,45°,90°,135°四个方向上差分图像,然后对差分图像分块计算每个图像子块的方差值,用最小的方差值与阈值比较决定其是否为模糊图像,并用图像的形态学操作去除检测时的噪点。实验表明算法对于高斯模糊和运动模糊图像都能较好的检测出图像的模糊区域。
This paper presents a new image partial blur detection based as four direction difference for partial blurry image.Four difference images are gotten by computing differential value of original image and their translation images translated recepectively in 0°,45°,90°,135°.Every differential image is partitioned into non-overlapping sub-blocks,which are signed blurry or not by comparing the least vari ance value to the threshold.morphologica operating is done f in order to eliminate noise dot.The results show that partial blur images can be effectively detected regadless of gaussian blur and motion blur.
作者
段新涛
宋黎明
孙印杰
DUAN Xin-tao,SONG Li-ming,SUN Yin-jie（School of Computer and Information Technology,Henan Normal University,Xinxiang 453000,China）
出处
《电脑知识与技术》
2012年第7期4725-4727,共3页
Computer Knowledge and Technology
关键词
局部模糊图像
四向差分
方差
高斯模糊
运动模糊
图像形态学操作
partial blur image
four direction difference
variance
gaussian blur
motion blur
image morphology operation
