基于热红外图像处理的保温墙体缺陷检测方法  

Defect Detection Method of Thermal Insulation Wall Based on Thermal Infrared Image Processing

作　　者：李翱翔 马志艳[1,2] LI Aoxiang;MA ZhiYan(School of Mechanical Engineering,Hubei Univ.of Tech.,Wuhan 430068,China;Research and Design Institute,Hubei Agricultural Machinery Engin.,Wuhan 430068,China)

机构地区：[1]湖北工业大学机械工程学院,湖北武汉430068 [2]湖北省农业机械工程研究设计院,湖北武汉430068

出　　处：《湖北工业大学学报》2023年第4期24-28,共5页Journal of Hubei University of Technology

基　　金：“十三五”国家重点研发计划(2018YFD0701002)。

摘　　要：冷库建筑围护结构的热工性能直接影响到冷库能耗。基于热红外图像的特征讨论保温围护结构缺陷检测方法,提出围护结构热工缺陷的判定指标:采用热红外图像信息熵特征对传统的Ostu算法的阈值判别函数进行改进;利用图像形态学处理与建立高斯混合模型的方式消除图像噪声;采用浸没模式的分水岭算法对图像进行分割。对冷库内外围护结构的图像采集与分析结果表明:形态学与高斯混合模型的建模增强了目标区域与背景的对比度,结合信息熵值所确定的分割阈值能准确提取目标信息,提高了图像分割边缘的准确率,分水岭算法的运用能快速且直观地描述各个区域的缺陷位置以及面积大小,缺陷分级标准评判了各种类型的缺陷程度,其实验结果验证了算法的正确性和有效性。With the development of urban energy-saving materials,the thermal performance of building envelopes directly affects the building energy consumption of the indoor environment.Therefore,more and more attention is paid to the defect detection of thermal insulation building walls at home and abroad.In this paper,based on the characteristics of thermal infrared images,a method for detecting defects in thermal insulation envelopes is proposed to determine the judgment indicators of thermal defects in the envelope:using thermal infrared image information entropy features to improve the threshold discriminant function of the traditional Ostu algorithm;using image morphology The method of learning and Gaussian filtering eliminates the image noise;the watershed algorithm of the immersion model is used to segment the image.The image acquisition and analysis results of the inner and outer protective structure of the cold storage show that the modeling of morphology and Gaussian mixture model enhances the contrast between the target area and the background,and the segmentation threshold determined by the information entropy value can accurately extract the target information and improve The accuracy of the edge of image segmentation,the application of the watershed algorithm can quickly and intuitively describe the defect location and area size of each area,and the defect classification simulates various types of defects.The experimental results verify the correctness and effectiveness of the algorithm.

关 键 词：围护结构 缺陷检测 图像分割 OSTU算法 分水岭算法 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]