Analysis of High Visibility Crosswalk Effectiveness in Improving New York State Safety
收藏DataCite Commons2020-07-15 更新2024-07-13 收录
下载链接:
https://dataverse.vtti.vt.edu/citation?persistentId=doi:10.15787/VTT1/I6YFO2
下载链接
链接失效反馈官方服务:
资源简介:
Project Description This study intends to provide new insight into the effectiveness of HVC markings in reducing risks of pedestrian vehicle crashes and conflicts at uncontrolled intersections. Data Request Scope 18 High Visibility Crosswalks (HVCs) were selected for this study, defined by GPS coordinates and roadway Link IDs. Data was generated in three phases. First, data was generated about all HVC crossings contained within SHRP 2 for a given period before and after the HVC was installed. Using this crossing data, specific HVC crossings were identified for additional analysis, and a second phase of data was generated containing front video and time series data, starting 15 seconds prior to crossing and 50 seconds after. Finally, a secondary extraction of data was requested from up to 500 events where a vehicle/pedestrian encounter was detected via video review. Some of the front videos and time series data from phases two and three are only available in a secure data enclave. Data Specification During the first phase, the following data was generated for each traversal for each of the 18 HVCs: • FILE_ID (unique identifier for each data file) • Date of trip (including day of month, month and year) • Binned time of the trip's entrance into the HVC road segment (local to the trip's location) • Participant ID • Participant age (as an age range) • Participant gender • Vehicle ID • Vehicle make • Vehicle model • Vehicle year • Trip exportability • Vehicle heading at the trip's entrance into the HVC road segment During the second phase, front video and time series data was extracted for specifically identified HVC crossings, providing (when available): • timestamp • FILE_ID • accel_x • accel_y • heading • latitude • longitude • speed_gps • speed_network • pedal_brake_state • pedal_gas_position • steering_wheel_position • video_frame • OBJECT_ID for lead vehicle • X_POS_PROCESSED for lead vehicle • Y_POS_PROCESSED for lead vehicle • IS_LEAD_VEHICLE • HEADWAY for lead vehicle In the final phase, the following data was extracted for up to 500 events:: • Eye glance location (coded from video) • Head rotation y • Head rotation y baseline • Head Confidence --
提供机构:
VTTI
创建时间:
2018-10-24



