In view of the current low efficiency of identifying high-consequence areas in long-distance pipelines,which requires a lot of manpower,material resources and time costs,this paper uses the target detection algorithm YOLOv5s,uses remote sensing images and drone flight image data,and combines high-consequence area identification specifications.Intelligent identification models for high-consequence areas are established for different types of high-consequence areas.Through evaluation,the model accuracy and recall rate have reached more than 90％,which can effectively identify high-consequence areas.Taking a high-consequence area of a domestic pipeline as an example to further verify the effectiveness of the model.The data shows that the information of the high-consequence area intelligently identified by the model is consistent with the traditional manual auxiliary identification information,and can meet the actual needs of automatic identification of high-consequence areas of long-distance pipelines.The actual needs provide timely and accurate data support for risk management of high-consequence areas,and open up new ideas for automated and intelligent management of high-consequence area identification.
long-distance pipelinehigh-consequence areaYOLOv5starget detectionintelligent recognition model
万方数据
维普
