Transmission Line Infrared Image Continuous Tube Overheating Detection Based on Artificial Intelligence Technology
This paper points out that with the development of modern society,artificial intelligence technology is continuously applied to the operation and maintenance of power grid equipment.In the past,in the infrared image of transmission lines,it was found that the main manual diagnosis experience of continuous tube overheating was low efficiency,subjective and the effect was not up to expectations.In this paper,artificial intelligence image analysis technology is used to automatically detect the overheating fault of continuous tube of transmission line in infrared image.Firstly,It detects the color bar in the infrared image of the transmission line,automatically establishes the mapping relationship between color and temperature;secondly,it uses the Yolov5 algorithm to automatically detect the continuous tube in the image,and then obtains the temperature of each pixel of the continuous tube according to the aforementioned temperature mapping algorithm,and automatically alarms when the temperature is found to be abnormal;finally,the algorithm is tested on a dataset containing 100 points.Experiments show that the proposed method can accurately detect the temperature abnormality of the continuous tube in the image and actively alarm,which has strong practical value.
万方数据
维普
