Maintaining social safety distance is an important means to prevent and control COVID-19. This paper proposes an application system based on YOLOv5 to build intelligent recognition of social safety distance. The system uses images captured by multiple cameras, recognizes pedestrians in the images based on YOLOv5 network, and calls the functions of the OpenCV open source library to calculate the real distance between pedestrians in three-dimensional space based on the two-dimensional images sampled by the cameras, and analyzes the acquired data on this basis to remind the people in the monitoring area to maintain a certain social safety distance. This system can be used for COVID-19 prevention and control, and made positive contributions to curbing the spread of COVID-19 virus.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Intelligent Recognition System of Social Safety Distance Based on YOLOv5

  • Huaizhong Zhu,
  • Yuguang Zhu,
  • Junyan Fan

摘要

Maintaining social safety distance is an important means to prevent and control COVID-19. This paper proposes an application system based on YOLOv5 to build intelligent recognition of social safety distance. The system uses images captured by multiple cameras, recognizes pedestrians in the images based on YOLOv5 network, and calls the functions of the OpenCV open source library to calculate the real distance between pedestrians in three-dimensional space based on the two-dimensional images sampled by the cameras, and analyzes the acquired data on this basis to remind the people in the monitoring area to maintain a certain social safety distance. This system can be used for COVID-19 prevention and control, and made positive contributions to curbing the spread of COVID-19 virus.