Visual Odometry

Extracted SIFT Features

This project develops a monocular continuous Visual Odometry (VO) pipeline designed for the purpose of vehicle localization. The work has been carried out as the project for the Visual Algorithms for Mobile Robotics course at the University of Zurich. Evaluation has been performed on three different datasets and as result, it is argued that the proposed pipeline (i) achieves local consistency in the estimated poses (ii) optimized through a continuous bundle adjustment approach, and (iii) can be further refined in terms of robustness and accuracy by employing inertial measurements.

The developed pipeline

Pascal Roth
Pascal Roth
Ph.D. in Robot Learning & Software Engineer

My research interests include deep learning, reinforcement learning, and computer vision to enhance robotic autonomy.