Crowd instability threshold monitoring method and system combining video monitoring and simulation

By combining video monitoring and simulation technology in high-density population monitoring, and using system entropy and Deepsort algorithms, precise monitoring and prevention of crowd instability stampede events is achieved, improving the efficiency and safety of emergency evacuation.

CN120014551AActive Publication Date: 2025-05-16NANJING TECH UNIV
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202510097637.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-01-21
Publication Date
2025-05-16
Estimated Expiration
2045-01-21

Smart Images

  • Figure CN120014551A_ABST
    Figure CN120014551A_ABST
Patent Text Reader

Abstract

The invention discloses a crowd instability threshold monitoring method and system combining video monitoring and simulation, and relates to the field of emergency evacuation. The method comprises the steps that a virtual environment is constructed, and high-density crowd data set pictures are adopted for training; combining the trained model with a Deepsort algorithm, loading an existing accident video, implementing perspective transformation, and extracting feature parameters of pedestrians; calculating pedestrian data in the detection frame of the selected person treading area to obtain speed density, and calculating system entropy data to obtain a threshold value of crowd instability; an Anylogic software program is used for building a scene, a region with the maximum density is locked, and a threshold value of crowd instability is input into a state diagram to achieve logic judgment; the real-time judgment of pedestrian walking logic is realized; and obtaining and monitoring the safety threshold of the critical number of people under the channel structures with different lengths and widths in the simulated scene.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Method for detecting personnel abnormal gathering behavior on the basis of video monitoring platform

    CN105930814A

  • Real-time crowd evacuation simulation method based on computer vision

    CN115830530A

  • Quantification and early warning analysis method for high-density crowd trampling risk of subway platform

    CN118230260A

  • Crowd evacuation system based on real time perception, simulation, and warning

    US11417106B1

  • A system based on ai and crowd dynamics invented to improve safety and efficiency while traveling in groups or clusters during crowded events like hajj and umrah

    WO2024013729A2