Research On Intelligent Video Detection Algorithm Of Icing And Foreign Matter On Transmission Lines

Transmission line is the main carrier of power transmission,its safe and stable operation is closely related to people's daily production and life.Icing and foreign matter attachment of transmission lines are two important factors that threaten its safety,and it needs to be accurately identified and judged.Aiming at the defects of poor accuracy,low intelligence level and insufficient anti-interference ability of existing methods for icing and foreign matter detection on transmission lines,a new method about icing and foreign matter detection based on intelligent image processing technology is proposed in this paper,focusing on improving the adaptability and accuracy of the algorithm and achieving the abnormal monitoring of transmission lines.Based on the characteristics of transmission line images taken by UAV and intelligent image processing technology,this paper studies the three aspects of image preprocessing,icing thickness recognition and foreign matter detection about transmission line images.Aiming at the characteristics of transmission line images,the common filtering and contrast enhancement methods for image preprocessing of transmission lines are studied.The advantages and disadvantages of several methods are compared.And new algorithms such as image classification processing based on color space variables and linear edge contrast enhancement based on wavelet analysis are proposed.A new method of icing thickness identification of transmission line based on binocular parallax image is proposed.The difference of gray value between foreground object and background object in parallax image is used to separate the foreground and background and a threshold selection method is proposed to calculated the binary threshold.Based on the principle of watershed algorithm combined with mathematical morphology to extract the complete contour of the icing transmission line.And the actual ice thickness is calculated according to the ratio of the number of pixels included in the icing contour and non-icing contourof the same column interval in the image.A new method of transmission line foreign body detection of the UAV image based on template matching and optical flow method is proposed.Based on gray morphology,a method for detecting the edge of transmission line which can filter out the background of the image is proposed.The concept that the moving speeds of corresponding pixels of objects with different distances from the camera in a specific shooting mode is difference is proposed.And we combine template matching and optical flow method to determine the location of foreign matter,and realize the accurate recognition of foreign matter through feature tracking.In this paper,the preprocessing method suitable for transmission line images is studied.A new image classification algorithm,a new edge contrast enhancement algorithm,an icing thickness recognition algorithm,and a foreign matter recognition algorithm are proposed.It can accurately and effectively realize icing thickness recognition and foreign matter detection,provids a new theoretical basis and technical means for intelligent monitoring of transmission lines,improves the intelligent monitoring system of the power system and promotes the process of smart grid construction.