| With the development of my country’s power grid system,the scale of the power grid is increasing day by day,and with it comes the thinking about how to ensure the security of the power grid.According to the statistics of the State Power Corporation,the number of short circuits caused by entangled foreign objects on overhead power lines and even grid system failures is increasing year by year.In order to solve the potential threats that various floating foreign objects may bring to power lines,foreign objects on power lines must be cleaned up in time.However,relying on traditional removal methods is not only inefficient,but also dangerous.In order to improve the efficiency of removing foreign objects and ensure the safety of the power grid,it is urgent to develop a device that can remove foreign objects without contact at a long distance.Therefore,this article will study a main controller used to control the obstacle removal equipment,using image processing technology to detect and identify the presence of foreign objects in the real-time power line image,accurately locate the area of the foreign object in the image,and track and identify the foreign object in the video sequence.Accurate location.The main research work of this paper is as follows:(1)According to the various hardware parts that make up the overhead power line foreign body laser obstacle removal instrument,the principal analysis is carried out according to the module,and the core components and PCB board design of the main controller are explained.According to the parameters of the camera and the pan/tilt used by this instrument,the camera is visually calibrated,and the image coordinate system is converted to the coordinate system of the pan/tilt.According to the transformation of the coordinate system,the accurate use of the video image to track the actual position of the foreign body is used for preprocessing.(2)Use the LSD(Line Segment Detector)line detection algorithm to detect and identify the position of the power line based on the image taken by the device camera,A method based on the least squares method of line fitting strategy to compensate the result of line fitting is proposed for complete detection of power lines containing foreign bodies.Use the gradient change of image pixels to detect the entanglement of foreign objects in the image,and use the minimum binding rectangle method to frame the foreign object target area based on the entanglement points and obtain the coordinate information.(3)The target area tracking technology based on Camshift algorithm is used to track foreign objects in the video sequence in real time.Research the tracking principle of the Camshift algorithm,the iterative processing mechanism of the Meanshift algorithm and the algorithm characteristics of the Kalman filter,and study an improved method of the Camshift algorithm fused with the Kalman filter.(4)According to the characteristics of the laser used by the instrument,design the cutting plan for the two kinds of power line wound foreign objects,and use the C++ to design and develop a supporting control software for the laser obstacle removal instrument.A scene simulation test is performed on the environment where the power line is wound with foreign objects to analyze the working conditions of the control software,and the foreign object laser obstacle removal instrument is controlled by the control software to verify the final cutting result. |
