Identification Of Bridge Crack Shape And Width Based On Unmanned Aerial Vehicle Imaging

As an important key node,the bridge bears the increasing traffic pressure.The technical status of the bridge is directly related to traffic and personal safety,thus it is very important to evaluate the technical status of the bridge.The biggest characterization of bridge disease is concrete cracking,in the past,the crack width was read by using a bridge inspection vehicle,it is difficult to describe the crack propagation.The original image of bridge surface is obtained by unmanned aerial vehicle,after analysis and processing,the information map of crack propagation is formed,which not only avoid artificial error but also improve the detection speed,and it is not restricted by the terrain,bridge type and bridge width,has important research value.This paper studies the problem of imaging angle correction of bridge detection by unmanned aerial vehicle,and put forward that:use three point laser to get three point distance,then deduce the average distance and relative angle of the measured surface,correct the pixel resolution obtained by distance method.Throughthe analysis of imaging characteristics of bridge detection by unmanned aerial vehicle,a combination filtering method based on morphology is proposed to preprocess image,enhance image contrast and protect edge information.By edge recognition of unmanned aerial vehicle's imaging,propose the Canny edge detection algorithm combined with Otsu,eliminate small noise by morphological criteria,get complete crack shape.The width of the fracture shape recognition algorithm to the least squares fitting method based on the crack of the center line,by fitting the curve tangent direction is calculated to crack width,and considering the special circumstances of the crack shape analysis and recognition.Obtaining the normal crack widthwith engineering significance provides data support for the technical evaluation of Bridges Based on unmanned aerial vehicle imaging.Finally,the paper carries on the experiment of real bridge and crack identification by UAV of Xiangtan xiangjiang river second bridge,compare the third party inspection agency data with width recognition results.The results show that the UAV imaging recognition has good accuracy and is superior to the crack extension information of the artificial bridge crack detection;it meets the actual needs of the project and has the high application value and potential.