During peak hours, the passenger flow of subway carriages increases sharply. In complex scenarios such as dense crowds and target occlusion, it is difficult to accurately identify each passenger, which can easily lead to missed or false detections. To this end, a subway carriage passenger flow detection method based on the improved YOLO algorithm is proposed. After analyzing the YOLOv8 model structure, the TFE module from ASF-YOLO was added to YOLOv8n. Combined with the spatiotemporal model, the characteristics of high passenger flow at stations and low passenger activity during driving, as well as the different flow characteristics of passengers in train doors and carriages, were considered. The multi frame detection results were fused to achieve accurate detection of passenger flow in subway carriages. Through experimental comparison, the average accuracy of the original YOLOv8n model is 57.0%, the improved model is 69.1%, and after multi frame fusion processing, it is 76.6%. The passenger flow information obtained based on this model supports multiple aspects such as passenger travel guidance, emergency rescue support, and railway operation control.
AguileraV, AllioS, BenezechV,et al.Using cell phone data to measure quality of service and passenger flows of Paris transit system[J].Transportation Research Part C: Emerging Technologies, 2014, 43(1):198-211.
[2]
ZhongG, WanX, ZhangJ,et al.Characterizing passenger flow for a transportation hub based on mobile phone data[J].IEEE Transactions on Intelligent Transportation Systems, 2016, 18(6): 1507-1518.
[3]
NiuQ, WuH F, GaoC Y, et al. Laser-based bidirectional pedestrian counting via height map guided regression and voting[J]. Signal, Image and Video Processing, 2017, 11(5): 897-904.
YuYan. Study on the pedestrian detection technology based on the information fusion of infrared CCD and laser measurement system[D]. Changchun: College of Electronic Science and Engineering, Jilin University, 2008.
WangLian-zhen, WangHui, LuoMeng-de. A detection system of taxi passenger flow based on pressure sensor[J] Technology & Economy in Areas of Communications, 2018, 20(5): 29-31.
[8]
SidenbladhH, BlackM J. Learning image statistics for Bayesian tracking[C]∥Proceedings Eighth IEEE International Conference on Computer Vision, Vancouver: Canada, 2001: 709-716.
FangChen-chen, ZhouJi-biao, DongSheng, et al. A method of forecast congestion of subway carriages based on BP neural network[J]. Journal of Transport Information and Safety, 2018,36(6): 47-53.
YangPu. Risk prediction and crowd evacuation of subway station congestion based on video image processing[D]. Beijing: School of Mechanical,Electronic and Control Engineering, Beijing Jiaotong University, 2018.
TanShi-lei, BieXiong-bo, LuGong-lin, et al. Real time detection of personnel mask wearing based on YOLOv5 network model[J]. Laser Journal, 2021, 42 (2): 147-150.
[15]
GaoC Q, LiP, ZhangY J, et al. People counting based on head detection combining Adaboost and CNN in crowded surveillance environment[J]. Neurocomputing, 2016(208): 108-116.
[16]
SongD, QiaoY, CorbettaA. Depth driven people counting using deep region proposal network[C]∥IEEE International Conference on Information and Automation(ICIA), Macau,China, 2017: 416-421.
[17]
AlqahtaniF, BanksJ, ChandranV,et al.Investigating the efficiency of a three-dimensional facial tracker using stereo camera arrangements[C]∥International Conference on Artificial Intelligence and Big Data,Chengdu,China, 2021: 463-468.
[18]
TranC K.Face recognition based on similarity feature-based selection and classification algorithms and wrapper model[J].EJournal Publishing, 2019(3):357-362.
[19]
ShenG, JamshidiF, DongD, et al. Metro pedestrian detection based on mask R-CNN and spatial-temporal feature[C]∥IEEE 3rd International Conference on Information Communication and Signal Processing (ICICSP),Shanghai,China, 2020: 173-178.
QiuFang, LiYu-feng, KongCai-hua. Real-time detection of passengers in subway carriages based on the fusion of YOLOv5m and attention mechanism[J]. Microprocessors, 2022, 43(6): 53-58.
[22]
LinS D, ZhuK X, ChenF,et al.Align-Yolact: a one-stage semantic segmentation network for real-time object detection[J].Journal of Ambient Intelligence and Humanized Computing, 2023,14(2):863-870.
ZhangXin, DongCheng-liang, WangXiao-chen, et al. Detection method of passenger flow density in train carriage based on improved YOLOv5s[J]. Railway Computer Application, 2022, 31(10): 10-16.
ZhouFei, GuoDu-du, WangYang, et al. Vehicle detection algorithm based on improved YOLOv8 in traffic surveillance[J]. Computer Engineering and Applications,2024,60(6): 110-120.