engleski

Computer Vision Based Indoor Navigation System for Ship Model

The subject of the bachelor thesis is sensory support for a navigation system. The final goal is the development of an algorithm on around stereo cameras to accurately estimate the internal velocities and orientation of the ship model. Except for the stereo camera, on the ship model is installed accelerometer, gyroscope, and magnetometer. To improve the accuracy of estimation, sensors are fused using Kalman and particle filter. Sensors are placed on the ship model, and the system is tested in a basin. Experiments were done in calm water conditions and wavy water conditions. Accuracy validation of estimations are done using an optical camera placed above the basin.

Computer Vision, Estimation, Kalman Filter, Particle Filter

nije evidentirano