基于改进动态阈值分割算法的光伏电池组件裂纹检测研究  

Research on Crack Detection of Photovoltaic Modules Based on Improved Dynamic Threshold Segmentation Algorithm

作　　者：曹茂深 黄勇 张思 孙先波 易金桥 梁树先 叶建聪 CAO Maoshen;HUANG Yong;ZHANG Si;SUN Xianbo;YI Jingqiao;LIANG Shuxian;YE Jiancong(School of Information and Engineering,Hubei Minzu University,Enshi 445000,China;Guangzhou Fengbiao Education Technology Company Limited,Guangzhou 510700,China)

机构地区：[1]湖北民族大学信息工程学院,湖北恩施445000 [2]广州风标教育技术股份有限公司,广州510700

出　　处：《湖北民族大学学报（自然科学版）》2020年第4期464-468,共5页Journal of Hubei Minzu University:Natural Science Edition

基　　金：国家自然科学基金项目(61661020);教育部产学合作协同育人项目(201802149044).

摘　　要：光伏电池组件是光伏发电系统的重要组成部分,它的破损直接影响到系统的输出功率和稳定工作,因此,对于光伏电池组件是否存在裂纹的检测是极为重要的环节.设计了一种针对光伏电池组件裂纹进行智能检测和标记的动态阈值分割算法,首先对无人机拍摄的图像进行预处理;然后通过不同宽高比的均值平滑来定位区域,利用平滑图和原图进行绝对做差,并排除干扰;接着对差值图进行阈值分割,小面积过滤除去杂质;最后通过闭运算拟合来标记裂纹,以此来确定有故障的光伏电池组件.该检测方法大大提高了光伏发电系统巡检工作的效率,具有较强的实用性和推广价值.Photovoltaic modules are an important part of the photovoltaic power generation system and their damage directly affects the output power of the system and stability operation.Therefore,the detection of cracks in the photovoltaic modules is an extremely important step.This paper designs a dynamic threshold segmentation algorithm for intelligent detection and marking of photovoltaic modules cracks.First,we preprocess the image taken by the drone,and then locate the area through the mean smoothing of different aspect ratios,using the smoothed image and the original image to make absolute difference and eliminate interference,and then perform threshold segmentation on the difference image to achieve a small area filtering and removing impurities,and finally mark cracks through closed calculation fitting to detect faulty photovoltaic modules,which greatly improves the efficiency of photovoltaic power generation inspection work and has strong practicability and promotional value.

关 键 词：光伏组件 裂纹检测 动态阈值分割 闭运算拟合 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]