In recent years drone racing has become popular, and for the first time in 2016 the ASL competed in an autonomous drone race. This race involved flying through a course made up of a series of obstacles and gates, with no human interaction.
Central to the system is the ability to accurately detect the objects within image and depth-camera data. The accuracy and speed of this detection is critically important to fast and reliable performance in the race. In this project the student will design a system for gate detection, implement it on our racing platform and test it within a small test track.
In recent years drone racing has become popular, and for the first time in 2016 the ASL competed in an autonomous drone race. This race involved flying through a course made up of a series of obstacles and gates, with no human interaction.
Central to the system is the ability to accurately detect the objects within image and depth-camera data. The accuracy and speed of this detection is critically important to fast and reliable performance in the race. In this project the student will design a system for gate detection, implement it on our racing platform and test it within a small test track.
During this project the following work packages will be undertaken:
- Familiarization with our existing gate detection algorithm, which includes GPU and non-linear optimization algorithms.
- Improvement of the speed and accuracy of the detection algorithm.
- Testing of techniques in simulation and on drone racing hardware.
During this project the following work packages will be undertaken: - Familiarization with our existing gate detection algorithm, which includes GPU and non-linear optimization algorithms. - Improvement of the speed and accuracy of the detection algorithm. - Testing of techniques in simulation and on drone racing hardware.
- Strong self-motivation and curiosity for solving challenging visual perception problems.
- Comfortable in C++ programming.
- A strong interest in algorithms, especially optimization and computer vision.
- Experience with Linux, ROS, and GPU programming are advantageous but not required.
- Strong self-motivation and curiosity for solving challenging visual perception problems. - Comfortable in C++ programming. - A strong interest in algorithms, especially optimization and computer vision. - Experience with Linux, ROS, and GPU programming are advantageous but not required.
alexander.millane@mavt.ethz.ch
Please include a CV, a transcript of records, and a short paragraph about your programming experience in your application.
alexander.millane@mavt.ethz.ch
Please include a CV, a transcript of records, and a short paragraph about your programming experience in your application.