大学生手机分心过街行为特征分析

李常兴，宋亚男，王诚心，韩浩然，李甜^*

山东交通学院交通与物流工程学院，山东 济南   250357

摘要：为分析以大学生为主的过街行人的手机分心行为特征，以济南市长清区大学城某交叉口为研究区域，设计试验实地调查并采集行人过街视频数据，将YOLOv8（you only look once version 8）目标检测算法与DeepSORT（deep simple online and realtime tracking）多目标追踪算法相结合，提取过街行人的手机使用情况、过街等待时间、启动延误时间、过街速度等行为变量，采用软件SPSS将提取的行为特征数据进行分类与统计分析。结果表明：未使用手机群体的过街等待时间集中在10~20 s，使用手机群体的过街等待时间集中在35~40 s，且等待时间波动较大，存在部分等待时间较长的异常情况，手机分心显著延长过街等待时间；未使用手机群体的启动延误时间集中在1~3 s，使用手机群体的启动延误时间集中在2~4 s，约59%的行人因手机分心导致启动延误时间超过3 s；未使用手机群体的过街平均速度为1.54 m/s，使用手机群体的过街平均速度为1.00 m/s；手机分心群体中出现闯红灯、斑马线外通行的人数占比分别为10%、53%。手机分心过街行为延长过街等待时间、启动延误时间及过街时间，增大违规通行与人车冲突风险，是慢行交通系统的安全隐患。

关键词：手机分心；过街等待时间；启动延误时间；过街速度；YOLOv8；DeepSORT

Analysis of street-crossing behavior characteristics under mobile phone distraction among university students

LI Changxing, SONG Yanan, WANG Chengxin, HAN Haoran, LI Tian^*

School of Transportation and Logistics Engineering, Shandong Jiaotong University, Jinan 250357, China

Abstract: To analyze the mobile phone distraction behavior of pedestrians, mainly university students, during street crossing, a crossroad in the University Town of Changqing District, Jinan City is selected as the study area. Field observations are conducted and video data of pedestrian crossings are collected. By integrating the YOLOv8 (you only look once version 8) target detection algorithm and the DeepSORT (deep simple online and realtime tracking) multi-object tracking algorithm, behavioral variables including mobile phone use, crossing waiting time, startup delay time, and crossing speed are extracted. The extracted behavioral data are then categorized and statistically analyzed using SPSS software. Results show that the waiting time of pedestrians not using mobile phones concentrates in 10-20 s, whereas that of pedestrians using mobile phones concentrates in 35-40 s, with greater fluctuation and occasional abnormally long waiting times. Mobile phone distraction significantly prolongs crossing waiting time. The startup delay time for non-users centers in 1-3 s, while for mobile phone users it centers in 2-4 s; about 59 % of pedestrians experience a startup delay exceeding 3 s due to phone distraction. The average crossing speed of nonusers is 1.54 m/s, whereas that of mobile phone users is only 1.00 m/s. Among the distracted group, the proportions of red-light running and crossing outside the crosswalk reach 10% and 53%, respectively. Mobile phone distraction extends crossing waiting time, startup delay, and overall crossing duration, increases the risk of traffic violations and vehiclepedestrian conflicts, and thus constitutes a safety hazard in slow-mode transportation systems.

Keywords: mobile phone distraction; crossing waiting time; startup delay time; crossing speed; YOLOv8; DeepSORT