基于模糊线段和链码相结合的关键点检测方法  

Key Point Detection Method Based on Combination of Fuzzy Line Segment and Chain Code

作　　者：戴玉静 李辉[2] 张静[2] Dai Yujing;

机构地区：[1]上海大学通信与信息工程学院,上海200444 [2]国电南瑞科技股份有限公司深圳分公司,广东深圳518000

出　　处：《工业控制计算机》2018年第10期77-78,81,共3页Industrial Control Computer

摘　　要：自然场景中对象的轮廓特征检测是计算机视觉领域中一个难点问题。提出一种基于模糊线段和链码相结合的关键点检测方法。通过十字模板提取图像中的线对象,线对象含有各个像素点坐标及其斜率等信息。将每个线对象用链码跟踪,分别将每个线对象进行分割,找到斜率发生较大变化的两个模糊线段之间的共有区域,在这个共有区域中找到曲率最大的点即为关键点。该算法应用于高压输电线路的自动巡检中,对无人机巡检所拍摄的输电线路图像进行实验。实验结果表明该方法可以在一定的光线变化范围和一定的背景复杂度内实现输电线路关键点和特征检测。A keypoint detection method based on fuzzy line segment and chain code is proposed in this paper.The line object in the image is extracted through the cross template.The line object contains information such as the coordinates of each pixel and its slope.Each line object is tracked with a chain code,and each line object is segmented to find the common area between the two fuzzy line segments where the slope changes greatly.In this common area,the point with the largest curvature is the key point.The algorithm is applied to the automatic inspection of high-voltage transmission lines.

关 键 词：模糊线段 链码跟踪 共有区域 关键点 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]