Research On Image Recognition Applied To Patrol Flight Control

In recent years,with the rapid development of China's national economy,the scale of the power grid has continued to expand,the length of transmission lines has increased rapidly,and problems have gradually emerged.First of all,there are very serious structural defects in power operation and maintenance.Secondly,traditional inspections rely on operation and maintenance personnel to walk on the ground and patrol the bases.The efficiency of inspections is extremely low,the maintenance and maintenance staff are safe and insecure,and the inspections are carried out.The quality of the inspection is low and the fault is not resolved in time.At present,the cases in which the UAV is applied to the transmission line inspection are controlled by the flying hand,which has great dependence on the flying hand and has great potential safety hazards.In order to solve the above problems,this paper proposes to apply image recognition to the research of flight control.In this paper,the machine vision method is used to judge the relative position of the drone and the patrol transmission line,and then the attitude of the drone is adjusted,so that the drone can independently complete the inspection of the transmission line.In terms of image preprocessing,a Wiener adaptive filter is designed to filter out the noise generated during aerial photography.Aiming at the problem of transmission line extraction,the canny algorithm is improved on the basis of comparing several edge detection algorithms,and the transmission line in the picture is extracted by combining the hough transform.Then a binocular vision ranging method combined with the improved SIFT algorithm is proposed to calculate the relative distance between the UAV and the transmission line.Finally,the solution method of the relative angle between the UAV and the transmission line is designed to complete the UAV and transmission.The position of the line is solved.Finally,based on the research of relevant theoretical knowledge,the overall design of the system and the software and hardware design of each part are completed.Through thecomparison of a large number of experimental results and simulation results,the feasibility of the proposed method is verified.reliability.Compared with the traditional transmission line inspection mode,the system greatly improves the inspection efficiency and safety,and can also be applied to bridge inspection,railway inspection,pipeline inspection,and independent landing.Good adaptability.