Dual-Camera Intersection Monitoring: Detection, Tracking, and Predictive Safety Alerts for Road Users
Bibliographic record
Abstract
Intelligent transportation systems (ITS) offer significant potential for enhancing traffic safety and efficiency through advanced sensing and driver assistance technologies. One of the major challenges for ITS is accurately assessing risks, particularly in interactions with unpredictable pedestrians in busy urban areas. This study addresses this challenge by developing a real-time camera-based monitoring system for proactive risk identification and mitigation at intersections, with a specific focus on ensuring the safety of drivers and pedestrians. Our research was conducted using data from two synchronized cameras that captured full HD video footage at the Gordon and Kortright intersection in Guelph City, Ontario, Canada. We collected and processed data from 40,132 frames per video, identifying and tracking road users using deep learning models. These models converted detected objects' coordinates to satellite and GPS coordinates, ensuring consistent tracking across both camera views. A novel multi-target multi-camera tracking algorithm was developed to maintain consistent object IDs across overlapping fields of view. This facilitated accurate trajectory prediction using deep learning models, incorporating features such as traffic light status, vehicle speed, and spatial interactions. The integration of these features significantly enhanced the model's predictive capabilities, enabling real-time risk assessments for pedestrians and vehicles. Key innovations include the use of computer vision techniques for traffic light status detection and the development of a binary 'Yield' feature to represent right-of-way laws. Additionally, our approach leverages satellite images to create a unified coordinate system for tracking, allowing seamless integration of additional cameras and sensors. The findings of this research demonstrate the feasibility and effectiveness of a camera-based monitoring system for real-time trajectory prediction and collision risk assessment. This system has the potential to improve urban mobility and road safety by reducing accidents and enhancing the safety of vulnerable road users.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".