基于图像Canny边缘检测和Hough变换算法的高温结构测量方法  

Method of High Temperature Structure Measurement Based on Canny Edge Detection and Hough Transformation Algorithm

作　　者：于合龙[1] 刘浩洋[1] 苏恒强[1] 王增辉[1] 

机构地区：[1]吉林农业大学信息技术学院,长春130118

出　　处：《吉林大学学报（理学版）》2014年第3期519-524,共6页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金(批准号:11372155);吉林省科技发展计划项目(批准号:201105068)

摘　　要：针对高温工程结构测量问题,将图像处理中的边缘特征检测技术应用到高温结构测量中,提出一种基于Canny边缘检测和Hough变换算法的高温结构测量方法,解决了高温结构测量的实际需求.该方法先对试件结构测量位置进行边缘检测和提取,并以分段直线形式确定待测边缘,再计算得到宽度等结构测量结果,并将其转换为材料力学性能参数.该方法充分利用了图像处理技术的优点,测量条件低且环境适应性强,同时采用分段检测的优化设计,测量精度高,稳定性好.According to the actual engineering requirements,the feature detection technology for image edge in image processing was applied to structure measurement at a high temperature.The hightemperature structural measurement method based on Canny edge detection and Hough transform algorithm was proposed.Based on the edge detection and extraction,the specimen structure is measured,which takes the form of piecewise linear representation.Then the width of the structure is obtained and it is converted into a mechanical performance parameter.This method takes advantages of image processing technology and needs low measurement conditions and has a good environmental adaptability.In addition,it adopts the optimization design of segment detection,with a high accuracy and good stability.

关 键 词：人工智能 结构测量 HOUGH变换 CANNY边缘检测 分段检测 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]