基于自组织定向的边缘优化方法

传统的边缘检测方法通常采用预先的平滑滤波操作排除噪声的干扰，这使得去除噪声的同时也损失了图像的细节。因此，滤波去噪比例和边缘检测精度一直是一对矛盾，本文针对这个问题提出了一种基于边缘定向的边缘优化策略。边缘具有一定的方向是它区别于杂乱无序的噪声的鲜明特征，利用这一特征，可以不需要平滑滤波操作并能很好地解决以上问题。文中还根据数字图像的特点提出了自组织定向这一获取图像边缘方向的有效方法。

Method of Edge Optimizing Based on Self-organizing Directing

The method of traditional edge detecting often adopts pre-filtering to eliminate the disturbing of the noise, which makes the details loss while getting rid of the noise. So, the proportion of filtering and the precision of edge detecting are a pair of contradictions all along. This paper presents a strategy of edge optimizing based on edge directing aiming at the question. It's a distinct difference compared with the noise that the edge has some direction, which needn't pre-filtering anymore and can make that problem solved primly. Meanwhile, we present a method of self-organizing directing which can obtain the direction of the edge effectively.