论文部分内容阅读
图像边缘检测的传统方法一般仅基于边缘邻域的一阶或二阶导数这一单一特征 ,并且对模糊边缘图像不敏感 ,针对此问题 ,本文提出基于多特征和模糊推理的边缘检测算法 .根据不同类别的边缘点附近灰度分布的特点定义多种边缘特征 ;利用不同的边缘特征确定模糊规则 ,用min max重心法模糊推理该点的边缘隶属度 ,实现边缘检测 .最后 ,文中给出应用本算法进行边缘检测的实例 ,与传统算法相比较 ,本算法检测边缘效果更好 ,并能较好地抑制噪声 .
The traditional method of image edge detection is generally based on the single feature of the first or second derivative of the edge neighborhood and is not sensitive to the fuzzy edge image. To solve this problem, an edge detection algorithm based on multi-feature and fuzzy inference is proposed in this paper. Different types of edge points near the gray distribution of the characteristics of the definition of a variety of edge features; use different edge features to determine the fuzzy rules min min center of gravity fuzzy reasoning the edge of the membership degree, to achieve edge detection.Finally, the paper gives the application The algorithm for edge detection example, compared with the traditional algorithm, the edge detection algorithm better, and can better suppress the noise.