| China is a vast country with a complex geographical environment,the transmission lines are exposed to the natural environment all year round.With the continuous improvement of people’s living standard,foreign bodies like kites,plastic bags and balloons will be hung on power lines more frequently.And the discharge distance can be shortened by hanging foreign bodies on the transmission line,It creates a threat to people and vehicles operating under the transmission line.However,the pictures taken by unmanned aerial vehicles are manually re-examined to find out whether there is any foreign body,which has a low efficiency and a high rate of false detection and mission.In view of this problem,this topic applies the image processing technology and the machine vision technology to solve this problem very well.The main contents of this research are as follows:1.The application status of UAV in transmission line inspection is analyzed.The status quo of transmission line extraction and foreign object detection research is analyzed.The methods of circuit extraction and foreign object detection are summarized.2.Aiming at the problem of uneven illumination and noise influence of UAV pictures,the algorithms of image gray-scale,histogram equalization,median filter and Gaussian filter are studied.3.Aiming at the problem of extraction of transmission lines in the presence of foreign objects,the transmission line extraction algorithm in the presence of foreign objects is designed: after image preprocessing,iterative threshold method is adopted to segment the image,mathematical morphology and connected region management are used to remove the influence of background after segmentation,and at last,probability hough transform is used to extract the line,and its validity is verified by experiment.4.Aiming at the problem of foreign object detection in transmission line,by summarizing the characteristics of foreign objects in the transmission line image,the algorithm of foreign body detection is designed: afterimage preprocessing,the foreign body is segmented by adaptive threshold,and the foreign body is detected by looking for the contour and according to the characteristics of the foreign body,the effectiveness of the method is verified by experiments.5.Though MFC combined with Open CV,an automatic foreign body detection system for transmission lines is designed and developed.The transmission lines are extracted,display whether foreign bodies are detected,improved UAV image detection efficiency and reduce the probability of missed detection. |
