Driving Simulator for Autonomous Marine Navigation: Development and Evaluation of Perception Algorithms
This project aims to extend the existing driving simulator at the MMP to include marine navigation capabilities, thereby enabling the testing of environment perception algorithms for both automotive and marine applications. The project will leverage the ROS/Gazebo platform and the NVIDIA Jetson Development Kit to develop and evaluate advanced algorithms for camera sensors. Initially, the algorithms will be prototyped and tested within the simulation environment, allowing for rapid iterations and refinement. Subsequently, the most promising algorithms will be validated on a real boat. The project will investigate the performance and robustness of the perception algorithms in diverse scenarios, contributing to the understanding of sensor-based perception systems in autonomous navigation.