Crowd counting using Yolov5 and KCF

Crowd detection has various applications nowadays. However, detecting humans in crowded circumstances is difficult because the features of different objects conflict, making cross-state impossible. Detectors overlapping zone may therefore overreact. In this paper, real-time people counting proposed using a model YOLOv5 (You Only Look Once) algorithm and KCF (kernel correlation filter) algorithm. The was used it considered one most accurate algorithms for real time. Despite high accuracy image, video or camera capturing, needs an increase speed. For reason, combined with tracking Where identifies to be tracked by KCF. trained on database people, system's reached 98%. speed system increased after adding