基于数学形态学和灰度矩的高温亚像素边缘检测

2019-03-14

作者：柯洁单位：合肥工业大学

摘要：提出了一种在高温条件下的试件边缘检测方法，能准确检测试件在升温过程中边缘的变化情况。对采集到的图像利用Canny算子和数学形态学进行像素级边缘粗定位，然后在灰度矩亚像素边缘检测算法基础上，增加对边缘判断条件的约束来检测亚像素级边缘，使得检测出的边缘更加真实可靠。经试验分析，该算法的分辨精度可达到0.1-0.15个像素，可实现高温条件下试件边缘的高精度非接触式测量。

关键词：边缘检测；数学形态学；灰度矩；亚像素

中图分类号：TG806；TH123.1文献标志码：ADOI：10.3969/j.issn.1000-7008.2019.03.033

Subpixel Edge Detection Method under High Temperature Based on

Mathematical Morphology and Gray Moment

Ke Jie

Abstract:An edge detection method for edge detection of specimens under high temperature is proposed，which can accurately detect the change of the edge of the specimen during the heating process.The Canny operator and mathematical morphology are used to locate the pixel level edge of the image.On the basis of the gray moment subpixel edge detection algorithm，the constraints of edge judgment condition are added，so that the subpixel edges can be detected，and the detected edges can be more true and reliable.The experimental analysis shows that the resolution accuracy of the algorithm can reach 0.1-0.15 pixels.This algorithm can achieve high accuracy and noncontact measurement of the edge of specimens under high temperature condition.

Keywords:edge detection；mathematical morphology；gray moment；subpixel