Learning objectives
The course aims to examine various perception-focused image processing algorithms for an "intelligent" vehicle.
The student will learn problems related to image processing for vehicles and in the laboratory he will implement algorithms that are normally used for safety and to allow autonomous navigation of a vehicle.
Prerequisites
A basic knowledge of image processing, C ++ programming, linear algebra and numerical calculation is required.
Course unit content
Within the course some image processing algorithms for intelligent vehicles will be presented and implemented: lane detection, obstacle detection, classification and tracking of obstacles and visual odometry. These algorithms can be considered as part of the group of "Advanced Driver Assistance Systems" and staring point for the development of fully autonomous intelligent vehicles.
Full programme
- Introduction
- Vehicle Issues
- Vehicle Sensors
- Data Fusion
- Sensor Calibration
- Visual Odometry
- Lane Detection
- Identifying obstacles
- FreeSpace, Occupacy Grid and Stixels
- Visual Self Localization and Visual SLAM
Bibliography
- - -
Teaching methods
The main topics of the course will benefit from laboratory activities and demonstations.
Assessment methods and criteria
Evaluation of laboratory activity, a written test and development of a research project.
Other information
- - -
