摘要
针对传统Canny算子在提取红外影像边缘时会出现毛刺、不光滑等现象,提出一种增强边缘细节的红外影像边缘检测方法。在Canny算子基础上,用主成分分析得到的主算子代替Sobel算子,同时用基于二阶偏导的非极大值抑制方法进行边缘细化,最后由"参考白"方法设置阈值提取最终边缘。经MATLAB仿真表明,相比Canny算子,所提方法能够提取更加平滑的边缘,尤其对弧形边缘。
As many cases such as skin needling and rough edge occur in edge extraction of infrared image by using traditional Canny operator, an improved edge detection algorithm is proposed. Firstly, principle operator obtained by PCA replaced Sobel operator,and non-max suppression method based on second derivatives are used to thin the edge. At last ＂reference white＂ method is used to extract final edge through setting threshold. Simulation results show that the proposed method can get smoother edge, especially for arch edge, compared with Canny operator.
出处
《激光与红外》
CAS
CSCD
北大核心
2015年第9期1129-1132,共4页
Laser & Infrared
基金
2014年浙江省大学生科技创新活动计划(新苗人才计划)项目(No.2014R407067)资助
作者简介
叶德周（1990-），男，在读研究生，研究方向为图像处理。
通讯作者：杨风健（1987-），男，在读研究生，研究方向为红外热像探测及图像处理。E-mail：yangfengjian2012@aliyun．com
