一种有效的图像边缘修正方法

AN EFFECTIVE METHOD FOR IMPRORING IMAGE EDGES

在实验的基础上,从视觉心理学出发,对边缘检测结果中的不理想情况进行分析和归纳,提出适合于边缘修正处理的链表型数据结构,并说明了边缘图像映射到数据结构中去的过程.根据目标域的线性、平行及封闭约束和图像边缘的长度约束以及边缘出错类型,形成修正规则与算法.最后给出了修正处理的实验结果.

Based on edge detection and psychology, six kinds of faults in the results of image edges are defined. secondly, a data structure is proposed and is suited for improving image edges.The procedune of mapping image edges into data strueture is also discussed. Then, according to the restrict of linear, parallel and closed features of object fields, and to the restrict of length of edge-image, as well as cases of faults in edge-image, a algorithm with improving rules is produced.