基于红外图像仪的建筑缺陷检测与识别研究

Research on Detection and recognition of building defects based on infrared imager

为了提高对建筑外窗气密性与缺陷的检测及识别,研究利用红外热成像仪采集建筑外窗的红外图像,并对图像进行预处理,最后通过检测程序获得外窗缺陷的检测结果。结果表明,研究模型的最优回归线与理想回归线趋势基本一致,其相关系数R分别为0.99703、0.99737、0.99745、0.99711。实际检测建筑外窗单位面积空气渗透量与程序检测的相对误差均值为4.038%,表明研究的检测程序具有较高的检测精度。此次研究对降低外窗耗能、提升建筑节能具有一定参考价值。

In order to improve the detection and recognition of the air tightness and defects of building exterior Windows,infrared thermal imager is used to collect the infrared images of building exterior Windows,and the images are preprocessed.Finally,the detection results of exterior window defects are obtained through the detection program.The results show that the trend of the optimal regression line and the ideal regression line of the research model are basically the same,and the correlation coefficients R are 0.99703,0.99737,0.99745 and 0.99711,respectively.The mean relative error between the actual air permeability per unit area and the program test is 4.038%,indicating that the proposed test program has high detection accuracy.This study has certain reference value for reducing energy consumption of exterior Windows and improving energy conservation of buildings.