基于OpenCV的变电站电力设备异物识别算法改进研究
Research on Improvement of Foreign Object Recognition Algorithm for Substation Power Equipment Based on OpenCV
张润泽1
作者信息
- 1. 国网陕西省电力有限公司超高压公司,陕西西安 710000
- 折叠
摘要
变电站智能巡视系统具备识别变电站高压电气设备上搭接鸟巢并发出告警的功能.但在实际应用中,摄像机拍摄的画面通常包含多台设备,导致智能巡视系统无法准确判断鸟巢位置,或将鸟巢搭接位置误判在其他设备上.为弥补异物识别算法无法准确定位的缺点,设计算法对摄像机拍摄的画面进行预处理,完成目标设备分离.对处理后的图像进行异物识别,将提高智能巡视系统的异物定位准确率.
Abstract
The substation intelligent inspection system has the function of identifying the overlapping bird's nest on the substation's high-voltage electrical equipment and sending out alarms.However,in practical applications,the images captured by the camera usually contain multiple devices,which makes the intelligent inspection system unable to accurately determine the position of the bird's nest,or misjudge the overlapping position of the bird's nest on other devices.In order to make up for the defect that the foreign object recognition algorithm can not locate accurately,the algorithm is designed to preprocess the pictures taken by the camera and complete the separation of target equipment.The foreign object recognition of the processed image will improve the accuracy of foreign object positioning of the intelligent inspection system.
关键词
变电站/智能巡视/图像处理/边缘检测Key words
substation/intelligent inspection/digital image processing/edge detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据