偏转(物理)
结构工程
分割
计算机科学
工程类
图像处理
数学
人工智能
光学
图像(数学)
物理
作者
Seyyed Hamed Farhang,Omid Rezaifar,Mohammad Kazem Sharbatdar,Alireza Ahmadyfard
标识
DOI:10.1061/(asce)cf.1943-5509.0001548
摘要
Deflection and displacement in building components such as beams and roofs occur under imposed loads or element weight during constructing or operation of buildings, and should be less than allowable values. The detection and measurement of deflection predominantly are conducted based on visual inspection, which is known to be time-consuming, costly, and error-prone. The authors used image processing techniques to detect the location of maximum deflection and measure its value. The perspective of color images was eliminated, and then they were preprocessed using spatial domain filters. Afterward, image segmentation, in combination with nonlinear regression, was used to calculate the equation of the deflection curve of an element. Finally, the location and amount of maximum deflection were determined by calculating the derivative of the deflection curve. The proposed method was validated through two experiments. The method estimated the amount of maximum deflection with high accuracy—the relative errors of the two tests were 6.6% and 0.2%.
科研通智能强力驱动
Strongly Powered by AbleSci AI