Development of road environment identification system using YOLO V5

The detection and recognition of road traffic environment is a key technology for smart cars, and it is also an important task of environmental perception in the vehicle’s automatic driving system. It is the basis for the subsequent path planning, decision-making and control of the vehicle. As the advancement of autonomous driving technology is becoming more and more mature, advanced driver assistance systems (ADAS) systems can complete partial semi-autonomous driving with human assistance, such as automatic cruise system (ACC), when fully autonomous driving technology has not yet been completed. The completion of fully autonomous driving technology is extremely important for road environment recognition. Therefore, this study uses the object detection technology YOLO V5 algorithm as the road environment identification. In the research, it is divided into two parts, the first part is the part of road environment image training, which mainly deals with image processing, establishing database and generating weight files for testing. The second part is the test system. When the specified image or image is input into the road detection system, data processing and identification are performed, and the identification result is labeled and output. The research results show that objects in the road environment can be marked through the YOLO V5 system as road environment recognition.