Research on the application of computer vision-based structural deformation monitoring

The continuous development of computer vision has brought vitality to the traditional discipline of civil engineering. Structural deformation monitoring based on digital image processing is a new non-contact measurement method that has come a long way since its introduction in the last century. Based on the research results of computer vision technology, the authors firstly perform a grey-scale transformation, filtering and noise reduction, and image enhancement on photographs collected in the laboratory and use edge detection algorithms for contour extraction. Structural displacements are obtained by combining higher-order polynomial fitting, and indirect structural strains using quadratic differential derivatives. In addition, deformation monitoring of a laboratory model was also completed using sensors. The results obtained were compared with the image processing method, which proved to be satisfactory for simple engineering applications.