Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Fusing Vision and Lidar Data to Improve the Position Estimate of a Quadrotor
The goal of this thesis is to develop an algorithm that will fuse depth information, obtained from a Lidar scanner, with odometry information, obtained from a Visual Inertial Odometry (VIO) system. Thus, an optimization method will jointly optimize discrepancies in the point cloud and the visual reprojection error. Your task is to start by implementing an offline version of this algorithm. We will then move to more advanced but realistic settings where this algorithm would run online or with data from a line scanner on a quadrotor.
Many robotic systems use Visual Odometry to estimate their position in the environment. However, this approach suffers from limitations such as a high dependence on good lighting conditions as well as position drift.
The goal of this thesis is to develop an algorithm that will fuse depth information, obtained from a Lidar scanner, with odometry information, obtained from a Visual Inertial Odometry (VIO) system. Thus, an optimization method will jointly optimize discrepancies in the point cloud and the visual reprojection error. Your task is to start by implementing an offline version of this algorithm. We will then move to more advanced but realistic settings where this algorithm would run online or with data from a line scanner.
The thesis is done in collaboration with Tinamu Labs, a young and dynamic startup with the focus of bringing robotic technology into the real world. You will have the opportunity to test your algorithm with Tinamu’s robotic system and collaborate with their engineers.
Many robotic systems use Visual Odometry to estimate their position in the environment. However, this approach suffers from limitations such as a high dependence on good lighting conditions as well as position drift. The goal of this thesis is to develop an algorithm that will fuse depth information, obtained from a Lidar scanner, with odometry information, obtained from a Visual Inertial Odometry (VIO) system. Thus, an optimization method will jointly optimize discrepancies in the point cloud and the visual reprojection error. Your task is to start by implementing an offline version of this algorithm. We will then move to more advanced but realistic settings where this algorithm would run online or with data from a line scanner. The thesis is done in collaboration with Tinamu Labs, a young and dynamic startup with the focus of bringing robotic technology into the real world. You will have the opportunity to test your algorithm with Tinamu’s robotic system and collaborate with their engineers.
In particular, the student will:
• Review literature on SLAM, 3D Reconstruction as well as fusion of vision and Lidar data
• Create a global optimization approach for the given task. Move to a more challenging setting with either computational constraints or degenerated data
• Verify the concept on captured datasets.
• Test and evaluate the algorithm in the real-world.
In particular, the student will: • Review literature on SLAM, 3D Reconstruction as well as fusion of vision and Lidar data • Create a global optimization approach for the given task. Move to a more challenging setting with either computational constraints or degenerated data • Verify the concept on captured datasets. • Test and evaluate the algorithm in the real-world.
We are looking for an independent student who:
• Is motivated to work with a real-world robotic system.
• Has basic knowledge and interest in SLAM and/or sensor fusion.
• Has interest and some background knowledge in Computer Vision.
• Has experience with Python or MATLAB (C++ is a plus).
•
Contact details:
Does this topic sound interesting to you? Please contact us! Please send a short cv and transcript of grades.
Alisa Rupenyan ralisa@ethz.ch
tobias@tinamu-labs.com
We are looking for an independent student who: • Is motivated to work with a real-world robotic system. • Has basic knowledge and interest in SLAM and/or sensor fusion. • Has interest and some background knowledge in Computer Vision. • Has experience with Python or MATLAB (C++ is a plus). • Contact details: Does this topic sound interesting to you? Please contact us! Please send a short cv and transcript of grades. Alisa Rupenyan ralisa@ethz.ch tobias@tinamu-labs.com