A binocular stereo visual servo system for bird repellent in substations

Bird damage has become one of the main threats to the safety of substations, and the accidents (short circuit, grounding failure and insulation flashover) caused by birds are increasing year by year. However, most existing devices and methods have little effect on intimidating birds in the substation, so they cannot meet the requirements for repelling birds in a high-efficiency and accurate way. In this paper, a binocular stereo visual servo system for bird repellent in substations is developed to detect, locate and repel birds in real-time. The system consists of a small processing computer, a rotatable Pan-tilt based on visual servo and the remote monitoring PC. In order to realize better real-time detection tasks, we improve the YOLOv3-tiny model by reducing the number of convolution kernels to reduce the parameters in the model and make the algorithm faster. And based on the bird detection results, we adopt binocular stereo vision and the coordinate conversion to obtain the distance and the angles of the bird relative to the laser. Experimental results in the actual working environment (the 220 kV substation) reveal that the proposed system could meet the real-time and accuracy requirements of detecting, locating and repelling the bird in practical substation application.