Background
Researchers at NTNU are currently developing an autonomous ferry designed for carrying pedestrians and cyclists between Ravnkloa and Brattøra. The development of a robust collision avoidance system for this ferry will benefit from a thorough analysis of traffic patterns in the channel. For this purpose a radar and a camera have been placed at Ravnkloa.
Scope
The goal of this project is to provide an analysis of the traffic in the Channel, by accessing data recorded by this system. The project will build on a previous specialization project / master thesis, where the choice was made to focus on analyzing the radar data, since these are easier to interpret than the camera data.
Proposed Tasks
- Implement a detector for extracting measurements of surface vessels in radar and/or camera data from the sensor station at Ravnkloa.
- Implement a tracking method in order to keep track of how different vessels in the Channel move.
- Generate statistics concerning how many vessels there are at different times of the day or the week, what are the most common routes/maneuvers, etc.
- Write report.
The project work aims to be extended into a master thesis for the spring of 2019. Depending on the interest of the students, there are several possibilities for this extension:
- More elaborate statistical analysis of the observations.
- Additional measurements with other sensors.
- Development of a sensor fusion system for collision avoidance that utilizes data from the stationary sensor station, together with other sensors onboard the ferry.
Supervisors
Main supervisor: Edmund Brekke.
Co-supervisor: Egil Eide.
Co-supervisor: Frank Lindseth.
