自适应阈值的边缘检测算法及其硬件实现  

Self-adaptive threshold edge detection and its implementation in hardware

作　　者：何文浩[1] 原魁[1] 邹伟[1] 

机构地区：[1]中国科学院自动化研究所高创中心,北京100080

出　　处：《系统工程与电子技术》2009年第1期233-237,共5页Systems Engineering and Electronics

基　　金：国家自然科学基金(60705026);国家"863"高技术计划(2006AA04Z242;2006AA04Z258)资助课题

摘　　要：传统Canny边缘检测算法的阈值需要人为设定,且计算量较大。为了克服传统Canny算法的这些缺点,提出了一种自适应阈值的边缘检测算法,并将其在硬件上实现。从图像的梯度幅值直方图分析出发,得出阈值的计算方法,自适应地动态生成阈值,克服了传统算法需要人为设定阈值的不足。介绍了自适应阈值的边缘检测算法在FPGA上实现的方法。实验结果表明,在FPGA上实现的自适应阈值边缘检测算法可以大大缩短算法所消耗的时间,具有实时性好、自适应性强的优点。The threshold of the traditional Canny algorithm needs to be set by manual, and its computation is complex. To overcome these shortcomings of the traditional Canny algorithm, A method of threshold-adaptable edge detection and its implementation in hardware are proposed. Based on the gradient histogram, the two thresholds can be found and the edge can be detected automatically. It is meanwhile recommended that an implementation of this edge detection is carried on in field programmable gate array （FPGA）. Experiments show that compared with the traditional canny algorithm, this method can detect edge automatically and needs much less time consumption.

关 键 词：边缘检测 CANNY算法 自适应 FPGA 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]