Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
3D Object Registration
The goal of this project is implementing and evaluating an object detection algorithm that is capable of dealing with clutter.
Keywords: ROS, C++, Robotics, Computer Vision
Aligning two 3d objects is a fundamental building block of many 3d vision algorithms. For simple cases involving good initial positions and point matches, this problem is addressed using the ICP algorithm.
However, many applications that require alignment may pose additional challenges such as a poor initial alignment or strong ambiguities.
The goal of this project is implementing (and potentially extending) an algorithmic alternative to ICP that is capable of handling object detection in a cluttered environment.
A successful project outcome will impact numerous areas of research at ASL that are not only limited to object detection, but also involve topics such as 3d reconstruction or SLAM.
Aligning two 3d objects is a fundamental building block of many 3d vision algorithms. For simple cases involving good initial positions and point matches, this problem is addressed using the ICP algorithm.
However, many applications that require alignment may pose additional challenges such as a poor initial alignment or strong ambiguities.
The goal of this project is implementing (and potentially extending) an algorithmic alternative to ICP that is capable of handling object detection in a cluttered environment.
A successful project outcome will impact numerous areas of research at ASL that are not only limited to object detection, but also involve topics such as 3d reconstruction or SLAM.
- Implementing and extending pointcloud alignment algorithms with a focus on cluttered sceneries.
- Evaluating the implemented algorithms by comparing them to other state-of-the-art approaches in simulations and using real data.
- Implementing and extending pointcloud alignment algorithms with a focus on cluttered sceneries. - Evaluating the implemented algorithms by comparing them to other state-of-the-art approaches in simulations and using real data.
- Excellent C++ (Python is a plus) coding skills (having written at least several thousand lines of code).
- A good understanding of computer vision, particularly 3d vision and ICP is desirable.
- Strong understanding of elementary probability and linear algebra.
- Previous exposure to ROS is a plus.
- Students from outside of D-MAVT (particularly, from D-INFK, D-ITET, D-PHYS, and D-MATH) are also highly encouraged to apply.
- Excellent C++ (Python is a plus) coding skills (having written at least several thousand lines of code). - A good understanding of computer vision, particularly 3d vision and ICP is desirable. - Strong understanding of elementary probability and linear algebra. - Previous exposure to ROS is a plus. - Students from outside of D-MAVT (particularly, from D-INFK, D-ITET, D-PHYS, and D-MATH) are also highly encouraged to apply.
If you are interested, please send your grade transcripts, CV, and a few sentences about your motivation/coding background to Fadri Furrer (fadri.furrer@mavt.ethz.ch) and Igor Gilitschenski (igilitschenski@ethz.ch).
If you are interested, please send your grade transcripts, CV, and a few sentences about your motivation/coding background to Fadri Furrer (fadri.furrer@mavt.ethz.ch) and Igor Gilitschenski (igilitschenski@ethz.ch).