摘要
角点是图像的一个重要局部特征,它决定了图像中目标的形状,因此在图像匹配、目标描述与识别及运动估计、目标跟踪等领域,角点提取都具有重要的意义。根据实现方法的不同可将角点检测分为基于边缘特征的角点检测、基于灰度图像的角点检测、基于二值图像的角点检测和数学形态学4类。详细阐述了这几种角点检测方法,并对不同方法逐一进行归纳分析,在最后指出了今后角点检测技术的研究方向和发展趋势。
Corner is a significant local feature of images,because the shape of the objective is decided by it,so it is very useful in object description, motion estimation and object tracking. In this paper, the corner detection algorithms are divided into four classes: boundary-based corner detection, gray level-based corner detection, binary image-based corner detection and mathematical morphology-based corner detection. This paper details the four corner detection methods ,and compared with each other. At last, it points out the future direction of research and development of corner detection.
出处
《机械工程与自动化》
2009年第1期198-200,共3页
Mechanical Engineering & Automation
关键词
角点检测
特征提取
边缘特征
灰度图像
corner detection
feature extraction
edge feature
gray image