Airfield pavement disease detection method and system

Abstract

The invention provides an airfield pavement disease detection method. The airfield pavement disease detection method comprises the steps of obtaining a trained YOLOv5 target detection model and a DDRNet semantic segmentation model; the trained YOLOv5 and DDRNet are optimized on the basis of a reasoning optimizer; the optimized YOLOv5, the optimized DDRNet and an image acceleration decoder are deployed on a movable inspection body; controlling the movable inspection body to run according to a preset running path, and controlling the image acquisition device to acquire an image; the image acceleration decoder acquires the acquired image in real time and decodes the acquired image; the decoded image is detected by the optimized YOLOv5, and a first detection result is obtained; and the optimized DDRNet performs contour information extraction on the image area with the disease based on the first detection result to obtain a second detection result. The invention also provides an airport pavement disease detection system. According to the invention, the detection result can be obtained in real time in the inspection process.

Description

technical field [0001] This application relates to the technical field of airport pavement disease detection, in particular to a method and system for detecting airport pavement disease. Background technique [0002] The airport pavement is an important facility of the airport. Under the influence of the climate, the airport pavement will gradually have cracks, incomplete marking lines and other diseases. The airport needs to pay attention to the dynamic changes of the above information at all times to eliminate potential safety hazards and ensure the safe operation of the airport. [0003] With the rapid growth of air traffic volume, airport pavement inspection has changed from the initial manual inspection method to the automatic inspection method. At present, radar detection technology and deep learning technology are the mainstream automatic detection technologies, but the price of lidar is relatively expensive, and the detection accuracy is difficult to meet the detecti...

AI Technical Summary

Problems solved by technology

At present, radar detection technology and deep learning technology are the mainstream automatic detection technologies, but the price of lidar is relatively expensive, and the detection accuracy is difficult to meet the detection requirements of small-sized cracks on the pavement; in the aspect of deep learning, target detection algorithm and semantic segmentation algorithm are used for processing Airport pavement images, but the target detection algorithm cannot extract the contour information of the cracks, so the severity of the cracks in the airport pavement cannot be judged; the contour information can be extracted using the semantic segmentation algorithm, but the real-time detection is poor, as disclosed in the patent document CN113111704A based on deep learning Method and system for detecting foreign matter on airport pavement

Images