基于BIM技术的建筑表面热工缺陷检测仿真  

Simulation of Thermal Defect Detection of Building Surface Based on BIM Technology

作　　者：栗丽[1] LI Li(Huanghe Science And Technology University School of Architecture And Engineering,Henan Zhengzhou,450005,China)

机构地区：[1]黄河科技学院建筑工程学院

出　　处：《计算机仿真》2019年第12期349-352,共4页Computer Simulation

摘　　要：针对传统的建筑表面热工缺陷检测方法存在检测所需时间较长、误检率较高等问题,提出基于BIM技术的建筑表面热工缺陷检测方法。结合概率分布函数与最大似然函数对建筑表面热工图像的噪声方差进行计算,自适应地调节双边滤波函数中灰度方差值,有效地抑制图像噪声;在图像去噪处理后,通过模糊聚类算法对建筑表面热工图像进行分割,分割出图像的缺陷区域;采用Harris角点检验算法对图像缺陷区域进行检测。实验结果表明,所提方法与传统方法相比,检测所需时间平均缩短了约4.9s,误检率降低了2个百分点。Traditional thermal defect detection time is too long and the false detection rate is high.Therefore,a method for detecting thermal defects of architecture surface based on BIM technology was proposed.At first,the probability distribution function was combined with the maximum likelihood function to calculate the noise variance of thermal image on the architecture surface.After that,the gray variance value in the bilateral filter function was adaptively adjusted,and thus to effectively suppress the image noise.After the noise reduction of image,the fuzzy clustering algorithm was used to segment the thermal image on architecture surface,so that the defect area of the image could be found.Finally,Harris corner detection algorithm was used to detect the image defect area.Simulation results show that the time required for the detection is shortened by about 4.9 s on average and the false detection rate is reduced by 2%.

关 键 词：双边滤波方法 中灰度方差值 模糊聚类算法 

分 类 号：TP391.4I[自动化与计算机技术—计算机应用技术]