测控技术2025，Vol.44Issue(6)：40-45,6.DOI:10.19708/j.ckjs.2025.06.301
基于双目视觉Gabor滤波器的光伏板清扫机器人目标识别
Target Recognition of Photovoltaic Panel Cleaning Robot Based on Binocular Vision Gabor Filter
邓随芳 1吴永宏 1周亚南 2宋建伟 2张魁 2李鹏2
作者信息
- 1. 国电投车师新能源(吐鲁番)有限公司,新疆吐鲁番 838000
- 2. 国电投河南工程运维有限公司,河南郑州 450000
- 折叠
摘要
Abstract
The photovoltaic panel cleaning robots are susceptible to factors such as changes in lighting and shadow interference during target recognition,making it difficult to accurately identify channel targets on the surface of photovoltaic panels.Therefore,a target recognition method of photovoltaic panel cleaning robot based on binocular vision Gabor filter is proposed.Based on the pose state and hand eye calibration type of the photo-voltaic panel cleaning robot,a mathematical model of robot motion is constructed.Combined with calibration er-ror correction,the hand eye calibration of the robot vision system is completed.Binocular vision technology is introduced to collect photovoltaic panel images,and a three-dimensional model of the photovoltaic panel is con-structed through stereo matching and disparity calculation.Then,Gabor filters are used to extract the texture features of the photovoltaic panel.Based on this,a target recognition classifier is constructed using support vec-tor machine(SVM)algorithm to output the centroid point coordinates of the cleaning target and determine its position,and the recognition of the photovoltaic panel cleaning target is achieved.The experimental results show that recognition misjudgment rate of the proposed algorithm for target recognition of photovoltaic panel cleaning robots is less than 0.25％,meeting the accuracy requirements and requiring less recognition time.关键词
双目视觉系统/Gabor滤波器/光伏板/清扫机器人/目标识别Key words
binocular vision system/Gabor filter/photovoltaic panels/cleaning robot/target recognition分类
计算机与自动化引用本文复制引用
邓随芳,吴永宏,周亚南,宋建伟,张魁,李鹏..基于双目视觉Gabor滤波器的光伏板清扫机器人目标识别[J].测控技术,2025,44(6):40-45,6.
