机电工程技术2024，Vol.53Issue(5)：162-166,5.DOI:10.3969/j.issn.1009-9492.2024.05.036
基于机器视觉的轨道线检测方法
Research on Track Line Detection Method Based on Machine Vision
摘要
Abstract
In order to address the issue of low accuracy and robustness in existing railway track detection and recognition algorithms,a series of effective image processing techniques are proposed to achieve track line detection.Firstly,the image is grayscale using the weighted average method,reducing the computational load and improving speed.Next,Gaussian filtering is applied to denoise the processed grayscale image,eliminating noise interference and enhancing subsequent processing accuracy.The Sobel operator edge detection algorithm is then utilized to extract edge information,effectively identifying edge features and accurately extracting track line edges.Finally,the extracted edge information is fitted using a cubic function to obtain a smooth envelope line.Experimental verification demonstrates that the proposed track line detection method significantly improves detection success rate,exhibiting excellent performance and high accuracy and robustness.关键词
机器视觉/边缘检测/灰度处理/弯轨拟合Key words
machine vision/edge detection/grayscale processing/rail bending fit分类
交通工程引用本文复制引用
赵鸿亮,郭佑民..基于机器视觉的轨道线检测方法[J].机电工程技术,2024,53(5):162-166,5.基金项目
国家自然科学基金资助项目(72061021) （72061021）
