This year SmartShip will focus on the detection and recognition of maritime signals and lights utilising camera, electronic navigation charts and other relevant data such as own vessel and surrounding vessel positions. This function is an important part of autonomous vessel systems required for safe and effective autonomous navigation.
As part of the situational awareness that is essential for autonomous vessel operations; the correct detection and recognition of maritime signals and lights is critical. There is a wide variety of information that is communicated through, for example, the use of navigational aids, lights on vessels and on shore, signs and other signals. For an autonomous system to utilise this information the signals must be detected, recognised and interpreted.
The team will develop computer vision and machine learning software for detection and recognition of various signals resulting in a translation of what the signals mean. Additional information such as orientation of a vessel may also be an output from the software.
Some of the disciplines covered by the project are:
- Computer vision including image processing and analysis
- Machine learning and Deep learning
- Managing and manipulating datasets
- Sensor fusion
- Software development
- Agile project management
Link to NRK video from the first year of SmartShip: