Pedestrian detection is the task of identifying and localizing individual people within images or video frames, typically by predicting bounding boxes or density maps around human figures in a scene. It serves as a foundational problem in crowd analysis and sensing research, enabling downstream tasks such as crowd counting, tracking, and behavior recognition, and its accuracy directly influences the reliability of systems deployed in surveillance, public safety, and smart environments. Key variants include single-pedestrian detection in uncrowded scenes, multi-pedestrian detection under occlusion or overlap, and density-based estimation approaches that shift from explicit localization to aggregate counting when individual detection becomes infeasible in highly dense crowds.

Source Papers

  • A survey of recent advances in CNN-based single image crowd counting and density estimation — A survey of recent advances in CNN-based single image crowd
  • Recent trends in crowd analysis: A review — Recent trends in crowd analysis: A review