Intelligent Inspection Method of Transmission Line Robot Based on Tower Image Registration and Object Tracking

To further enhance the intelligence of the transmission line inspection robot, this paper proposes an intelligent inspection method based on the registration of the tower image and the tracking of inspection objects. During the inspection process of the same type of towers, this method completes the mapping of all inspection points from the template through the matching and registration of SURF feature descriptors in the tower image. Then, according to the initial position of the inspection point, the real-time update of the inspection target position is realized by the object tracking algorithm based on correlation filtering, so as to automatically adjust the angle of the PTZ to complete the inspection shooting. The results show that the registration method based on the image of the tower effectively realizes the reuse of the inspection point template. When the IoU threshold is 0.6, the accuracy of the inspection point mapping reaches 93.7%, and the qualified rate of the automatically collected photos of key components reaches 92.2%, which can meet the requirements of fine line inspection.