Automatic 3D Video Analytics for Traffic and Road Condition Assessment
Measurement of highway traffic and road conditions is important for planning, management, incident detection and notification, all of which are crucial to the efficiency and safety of our vehicle transportation systems. With the availability of highway camera networks, such as MTO’s COMPASS system, there is opportunity to use continuous video data to monitor highway traffic and assess road conditions. Human experts can analyze video but this is costly and cannot be done thoroughly on a continuous basis.
There is therefore a requirement for reliable, sustainable, fully automatic traffic analytics systems that can deliver accurate traffic metrics on a continuous basis. The goal of this project is to help meet this requirement through the research and development of fully-automated traffic analysis hardware and software, demonstrated, evaluated and compared with commercial systems on MTO highways.
To find out more:
- Almazen, E.J., Qian, Y. & Elder, J.H. (2016) Road segmentation for classification of road weather conditions. Proceedings of the 4th Workshop on Computer Vision for Road Scene Understanding and Autonomous Driving, ECCV 2016, Springer, 96-108.
- Corral-Soto, E. and Elder, J.H. (2014) Automatic single-view calibration and rectification from parallel planar curves. In D. Fleet, T. Pajdla, B. Schiele, and T. Tuytelaars, editors, ECCV (IV), volume 8692 of Lecture Notes in Computer Science, Springer, Berlin, 813–827.