Back to datasets
Dataset assetOpen Source CommunityVehicle Trajectory AnalysisTraffic Surveillance
Ubiquitous-Traffic-Eye
This dataset contains vehicle trajectory data from multiple locations, extracted from aerial video. Data accuracy is ensured through manual labor. The dataset includes video files, raw trajectory data, and data format descriptions, recording detailed parameters such as vehicle count, position coordinates, lane number, vehicle length, width, and speed. The temporal precision is 0.1 seconds and spatial precision is 0.01 meters.
Source
github
Created
Apr 8, 2022
Updated
May 20, 2024
Signals
385 views
Availability
Linked source ready
Overview
Dataset description and usage context
Dataset Overview
Dataset Name
Ubiquitous-Traffic-Eye
Dataset Content
- Trajectory Data: vehicle trajectory data from multiple locations, sourced from aerial video.
- Data Precision: temporal precision of 0.1 seconds, spatial precision of 0.01 meters.
- Included Parameters: vehicle ID, position coordinates, lane number, vehicle length, vehicle width, speed, time interval, spatial interval, acceleration/deceleration, etc.
Downloadable Content
- Video Files: contain object detection bounding boxes.
- Raw Trajectory Data: detailed vehicle trajectory records.
- Data Format Description: provides detailed specifications of the data format.
Dataset Distribution
- SQM: includes SQM-W-1, SQM-W-2, SQM-N-4.
- YTDJ: includes only YTDJ-3.
- CKQ: includes only SQM-N-4.
- XAM-N: includes XAM-N-5, XAM-N-6.
- RML: includes only RML-7.
- PKDD: includes only PKDD-8.
- XAM-S: includes only XAM-S-9.
Dataset Characteristics
- Data Source: accuracy ensured through manual labor.
- Data Diversity: collected across various locations and times, covering dynamic traffic conditions.
Need downstream help?
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.