Visual Exploration of Road Traffic for Sensor Data Fusion
Evaluating the reliability of sensors represents an important basis for designing stand-alone modules that should provide real-time traffic monitoring. In this project we aim to assess reliability of infrared and radar sensors for detecting persons and vehicles in daily traffic. To enable such observation we also provide constant video surveillance data, which represents a ground truth for comparison with the sensor data (e.g., the radar detected a car approaching at a given speed). The goal of the project is to develop a visual exploration framework that would allow to navigate and process interesting video frames (e.g., via machine learning and / or image processing techniques) while integrating the sensor metadata into the exploration process.
The project is part of Decca Lighting activities http://www.deccalighting.com/
Design and implement visual exploration tool to cross-validate significant video frames with the sensor data (Infrared and radar).
Skills: C++, Java or python, Image processing and basic machine learning
- Daniel, Gareth, and Min Chen. “Video visualization.” Proceedings of the 14th IEEE Visualization 2003 (VIS’03). IEEE Computer Society, 2003.
- INF219 & INF319
- Master thesis project (The project can be expanded)
For more information please contact Julius Parulek (firstname.lastname@example.org).