The Identification of Device Status Indicator in Hydropower Station Based on YOLOv5s
Ningjun Dang, Hexiang Huang, Yanni Kang, Shuoxun Li, Xinwei Zhou · 2023
In order to timely detect abnormal status indicators of hydropower stations when no one is on duty, it becomes very important to develop a real-time detection system for achieving remote alarms. To achieve this goal, this paper studies the detection and identification of facility status indicators by utilizing the surveillance camera in hydropower station. We firstly collect many camera based images from various station indicators and construct the training datasets. Then we choose to use the lightweight YOLOv5s model for detecting images captured by the camera. Finally, the detection effectiveness is evaluated by the average accuracy. The test results show that the model can effectively identify changes of device status indicators in hydropower station.