Pedestrian flow monitoring system based on Deep learning pipeline

This paper designs a pedestrian flow monitoring system, which uses a variety of deep learning technologies (including target detection, target tracking, video classification, etc.). Three different schemes are designed according to different Pedestrian flow scenarios. A system composed of deep learning pipeline, data visualization and database is built to monitor the pedestrian flow in crowded public places, helping managers to better collect and manage the pedestrian flow data.