No abstract

Based on the current situation of passenger flow in Guangzhou Metro, this paper summarizes the shortcomings and possible risks of passenger transport organization. And after analyzing the current main methods of monitoring crowd density, and the crowd density detection algorithm in video surveillance and subsequent video processing, this paper proposes a crowd density classification method based on frame difference method with its application direction, which optimizes the existing subway station crowd control measures as well as helps subway station operators accurately identify the stage of crowd density in real time and take crowd control measures in time.