Research On New Bridge Inspection Technology Based On UAV Photogrammetry

In the face of China's huge bridge group,the traditional bridge inspection methods show low efficiency,low degree of automation,high inspection cost and other shortcomings,and gradually become the pain point of the industry.In recent years,UAV has become more and more active in various industries,and its most prominent advantages are flexibility,low cost,high degree of automation and excellent information acquisition ability.As an advanced tool,it can better replace human eyes to observe the appearance of the bridge with high-definition lens.The task content of bridge regular inspection is mostly completed by visual inspection,which is in good agreement with the function of UAV to obtain image information.To a certain extent,it can initially deal with the content of bridge regular inspection task.Taking the above situation as the basic research status,this paper proposes a bridge inspection technology method based on UAV photogrammetry.Taking the concrete bridge as the research object and to be inspected,on the premise of photogrammetry as the theoretical basis,the UAV is used to obtain a large number of image information of the bridge to be inspected,and the computer image processing technology is used to process the relevant images,identify the disease and quantify the characteristic information of the disease,establish the three-dimensional real scene model of the bridge,and realize the location of the damage and disease.The process of bridge inspection is automatic and efficient,the inspection results are presented in the form of model,and the disease information is clear.The integration of UAV and bridge inspection industry is preliminarily realized,which brings certain convenience for bridge inspection.In order to complete the above research,the main work of this paper is as follows:(1)This paper studies and expounds the theory of photogrammetry.Starting from the concept of photogrammetry,the definition of photogrammetry theory,the important parameters and the geometric relationship between them are clarified.This paper studies the common coordinate system of photogrammetry,the exterior orientation elements of photograph,the transformation of space rectangular coordinate and the collinear condition equation.In this paper,three kinds of photogrammetric analytical processing methods are studied,which lay a theoretical foundation for the later research.(2)This paper introduces three common path planning algorithms for UAV,analyzes their advantages and disadvantages,and proposes a new path planning method for UAV Based on point cloud model for bridge inspection.Firstly,get the point cloud model of the bridge,find the characteristic points of the bridge components,use the geometric information of the bridge components expressed by the characteristic points,and write the UAV path planning algorithm in MATLAB.The algorithm is used to calculate the waypoint position information of UAV during automatic flight,and the attitude parameters such as pitch angle and yaw angle of UAV PTZ are calculated according to the UAV flight mission,which are incorporated into the KML file that can be recognized by UAV,so as to realize the path planning of UAV and further improve the inspection efficiency of UAV.(3)With the help of experiments,this paper studies the measurement and location of apparent damage recognition of bridge image,and realizes image noise reduction by three image preprocessing methods: image graying,image filtering and image enhancement.Then,the advantages and disadvantages of first-order and second-order differential operators in image edge detection are studied.After comparison,it is concluded that canny second-order differential operator is more suitable for image edge detection,and the neighborhood state value deletion method is proposed to delete the redundant noise after edge detection.In the bridge inspection,the most important thing is to quantify the cracks.On the basis of edge detection,the area,length and width information of cracks are calculated,and compared with the actual crack feature information to verify the accuracy of the calculated feature information.Finally,the establishment method of 3D model,fracture location method and modeling accuracy of the model are studied.(4)The UAV path planning,image processing,3D modeling and damage location in the model are studied.The bridge inspection technology based on UAV photogrammetry is applied to the actual bridge,and the apparent diseases of the bridge are obtained,and good results are achieved.