Research On The Monitoring Of Concrete Dam Surface Crack Based On UAV Image

Dam is an important infrastructure of China’s national economy,which plays an important part in flood control,power generation and irrigation.However,with the extension of its service time,the dam will appear various structural damage,which will affect the safe operation of the dam.Among them,cracks are the most common damage problems of concrete dams.However,there are few monitoring methods for concrete dam cracks,and the conventional methods are difficult to implement and have poor operability.Therefore,it is urgent to develop a new technology to monitor the apparent cracks of concrete dams.In this paper,according to the current deficiencies of concrete dam safety monitoring technology,combined with machine vision method and UAV technology,an apparent crack monitoring system for concrete dams based on UAV images was established,and the imagesstitching & crack-detection software based on UAV images was developed,and the software is applied to crack detection in a reservoir engineering,achieved good results.Image mosaic includes three key steps: image preprocessing,image registration and image fusion.For image preprocessing,the process of image space transformation is studied,and the solution of camera parameters and coordinate transformation matrix between images is analyzed by using the transformation relation between world coordinate system and pixel coordinate system.The research compares the commonly used image enhancement algorithm and image denoising algorithm to improve the image quality of UAV.For image registration,this paper analyzes and compares the effect of SIFT operator and Harris operator to extract feature points,applies RANSAC algorithm to select feature points;analyzes common direct average fusion method,weighted average method and multi-band fusion method,and an image fusion method of concrete crack based on optimal seam lines search is presented.For crack segmentation,this paper analyzes and compares the principles of Roberts operator,Sobel operator,Prewitt operator,Canny operator and bimodal method,automatic iteration method and maximum inter class variance method,and their application effect in crack segmentation;PSO algorithm is introduced into threshold segmentation,and the inter class variance function is taken as the objective function,and a good improvement effect is achieved.Finally,based on the image preprocessing,image mosaic and crack recognition algorithm studied in this paper,an images-stitching & crack-detection software based on UAV images is developed based on MATLAB language,and a reservoir project was used as the test site for software application.The results show that the software developed in this paper can realize the functions of UAV images preprocessing,stitching and crack detection,and have practical engineering application value.