基于尺度空间理论的角点检测  

Corner Detection Algorithm based on Theroy of Scale-space

作　　者：邓传斌[1] 郭雷[1] 李维[2] 

机构地区：[1]西北工业大学自动化学院,西安710129 [2]西安工业大学电子信息工程学院,西安710032

出　　处：《火力与指挥控制》2010年第11期26-28,共3页Fire Control & Command Control

基　　金：国家自然科学基金重大研究计划(90205019);国家自然科学基金资助项目(60802084)

摘　　要：传统的Harris角点检测是一种经典的角点检测算法,但不具有尺度不变性。研究了一种基于尺度空间理论的Harris角点检测方法。建立Harris的尺度空间函数表示。利用多尺度分析的思想,使在小尺度上的角点精确定位和大尺度上的角点去伪存真。从而克服了传统的Harris角点检测可能存在的角点信息丢失、角点位置偏移和易受噪声影响等问题。试验结果表明,该算法明显地提高了图像角点检测性能。Traditional Harris corner detection is a classical algorithm, but has not the scale invariant property. In this paper, an improved method of Harris corner based on the scale-space theory was described,and a scale space representation of Harris corner was established. Point corner located accurately on a small scale, eliminated the false and retained the true on a large scale by using the multi-scale idea. Then, traditional Harris detector with missing significant corners and detecting false corners due to noise was overcame. Experimental results show that the proposed rnethed is more efficient than traditional methed in detecting the corners.

关 键 词：HARRIS角点检测 角点检测 尺度不变性 尺度空间理论 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]