Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Robust state estimation and visual tracking
Work on a VIO /SLAM algorithm for robust state estimation and tracking for industrial robotic applications.
Keywords: SLAM, Visual Odometry, Tracking
Visual Odometry (VO) and Simultaneous Localization and Mapping (SLAM) have gained a lots attention from industry in recent years. Several obstacles, such as robustness, accuracy, etc., are still in the way when the VO, Visual SLAM (V-SLAM) technologies are applied in real industry applications. Joint with an industry company, we investigate the robustness of an on-line tracking approach on body-worn devices with fast movements such as running, jumping, roping and turns. The final application is supposed to be run with multiple maps and under dynamic lighting conditions ranging from direct sun-light to almost no-light, e.g. in a dark room.
Requirements: Understanding of VIO and SLAM, Computer Vision and Programming experience in C++.
Visual Odometry (VO) and Simultaneous Localization and Mapping (SLAM) have gained a lots attention from industry in recent years. Several obstacles, such as robustness, accuracy, etc., are still in the way when the VO, Visual SLAM (V-SLAM) technologies are applied in real industry applications. Joint with an industry company, we investigate the robustness of an on-line tracking approach on body-worn devices with fast movements such as running, jumping, roping and turns. The final application is supposed to be run with multiple maps and under dynamic lighting conditions ranging from direct sun-light to almost no-light, e.g. in a dark room.
Requirements: Understanding of VIO and SLAM, Computer Vision and Programming experience in C++.
Development and performance evaluation of a VIO / V-SLAM algorithm for state estimation and tracking. The obstacles and difficulties in challenging scenarios as fast movements and tracking in dynamic lighting conditions will be identified and solutions will be proposed and evaluated accordingly.
Development and performance evaluation of a VIO / V-SLAM algorithm for state estimation and tracking. The obstacles and difficulties in challenging scenarios as fast movements and tracking in dynamic lighting conditions will be identified and solutions will be proposed and evaluated accordingly.