一种利用嵌入式技术的SF_(6)仪表图像识别方法

Image Recognition Method of SF_(6) Instrument Using Embedded Technology

SF_(6)作为良好的气体绝缘体,被广泛应用于变电站电气设备的气体绝缘。变电站内多为大型设备,SF_(6)压力仪表安装的位置超出人眼可观测范围,无法通过人工观察准确获得仪表指针位置及读数。因此,设计一种基于小型智能相机镜头的嵌入式设备,实现对SF_(6)仪表拍摄和指针识别十分必要。首先,使用嵌入式终端主机控制智能相机镜头对SF_(6)仪表进行拍摄;随后,智能相机镜头通过WIFI网络把拍摄照片传送至嵌入式手持终端主机;然后,手持终端对拍摄照片执行图像处理任务;最后,对处理后的照片进行图像识别,最终确定仪表指针位置。在某220 kV变电站进行试验,结果表明,SF_(6)仪表指针识别准确率达98%以上,达到应用预期效果。

As a good gas insulator,SF_(6) is widely used in gas insulation of electrical equipment in substations.In substations,most of the equipment is large and the SF_(6) pressure gauge is installed in a position beyond the observable range of the human eye,making it impossible to obtain accurate gauge pointer positions and readings through manual observation.Therefore,it is necessary to design an embedded device based on a small smart camera lens to enable photography and pointer recognition of SF_(6) meters.Firstly,the SF_(6) meter is photographed using the embedded terminal host to control the smart camera lens.The smart camera lens then transmits the captured photos to the embedded handheld terminal host via a WIFI network.The handheld terminal then performs image processing tasks on the captured photos.Finally,image recognition is performed on the processed photos to finally determine the position of the instrument pointer.The test is carried out in a 220 kV substation and the results show that the SF_(6) meter pointer identification rate reached over 98%,achieving the expected effect of the application.