SAE-GFR Evaluation and improvement of racetrack detection with cameras
Camera image detection is currently our primary method of detecting the race track. There are multiple projects in the camera pipeline this year.
The first will be to retrain our YOLOv4 neural network on the public dataset created by the teams of Formula Student. Additionally, they will develop and implement a process to relabel our own dataset to crowdsource image labeling across the team.
The second project will be to automate the pipeline for intrinsic and extrinsic calibration. Using open source backends, an interface for collecting, processing, and saving calibration data and results will be developed, and a documented process for executing calibrations with a focus on repeatability
The third project will develop a process to quantitatively evaluate all of the team’s depth estimation algorithms. Another modern depth estimation algorithm will be implemented, and the real-world accuracy and performance will be evaluated with the best being implemented.
Objectives
- Fall
- Background research of previous projects and relevant papers completed.
- Github and development environment set up.
- ROS tutorial and assignment completed.
- Design changes conceptualized.
- Practice for participation in rules quiz.
- Documentation for any changes made during Fall.
- Winter
- Final documentation for code completed - details on wiki as well.
- Testing and benchmarking completed and documented.
- UML Diagram detailing system.
- Final edits to the code by May.
Motivations
Camera is one of two ways that the autonomous vehicle interacts with the world. Working on and improving the camera pipeline has the potential to directly improve the performance of an autonomous race car.
Qualifications
Minimum Qualifications:
None Listed Preferred Qualifications:
None Listed
Details
Project Partner:
Marcus Wheeler
NDA/IPA:No Agreement Required
Number Groups:1
Project Status:Accepting Applicants
Website:https://www.global-formula-racing.com/en/