Abstract:The rapid detection of road diseases is crucial to ensure the safety and reliable operation of roads. The ground penetrating radar (GPR) technology has been widely applied in road disease detection because of its speed, non-destructiveness, high resolution and other features. However, the previous radar image processing and interpretation were mainly relied on the subjective experience of personnel, which leads to the misjudgments and missed detections. To solve this problem, an intelligent road disease recognition system has been developed by studying the image recognition method based on YOLO algorithm and combined with the deep learning technology. This system can automatically extract the characteristics of various diseases from GPR images and achieve the efficient and intelligent recognition. And the accuracy of the recognition results is ensured through the borehole verification to effectively prevent the sudden road collapses and improve the road safety and reliability.
