Transmission Line Abnormal Target Detection Based on Machine Learning Yolo V3

The power video surveillance system must be able to quickly detect abnormal targets of power lines, detect and alarm suspicious external intrusions, focusing on the detection of off-line operations of large construction vehicles such as cranes and excavators, so as to prevent the damage or failure of power lines caused by external intrusions of power lines. Based on machine learning Yolo V3 algorithm, this paper trains and learns the routine intrusion targets of power lines, realizes the detection of intrusion targets of power lines in real time, discovers abnormalities in time and protects the safety and normal operation of power lines.