Research on intelligent inspection technologies of expressway maintenance based on edge computing

Smart inspections of expressway maintenance are collected on road data through vehicle cameras. The vehicle processing equipment combines edge computing to preprocess the collected image to reduce the pressure of data transmission background data operations. The back-end combination of image recognition technology and deep learning theory to filter the collected pavement information, and finally realize road disease recognition. This paper is based on technologies such as images, image enhancement, and data expansion. Through the research of the FASTER RCNN and YOLO-V5 algorithms, the UNET and YOLO-V5 are finally used for object segmentation and recognition.