Students are invited to take part in Technion's Formula team and develop a driverless formula vehicle.
Students are invited to take part in Technion’s Formula team and develop a driveress formula vehicle. See demo movie and some details below.
Formula Technion Driverless team – High Speed Image processing project
High speed object detection via Image processing and LiDAR technology
for an autonomous race car as part of the Formula Technion Driverless team
The Formula Student Automotive Society (FSAE) is an international series of racing competitions and a project challenging engineering students to design and build a one-of-a-kind racecar to compete in international competitions against top-tier universities. This year, for the second time ever in the world, an autonomous race car competition is being held and we are planning on being a top competitor in the race. Based on the 2016 combustion engine Formula race car, an autonomous vehicle is being designed and brought to life and you can be a major part of it.
2. Project Specification:
As part of the track detection and foresight needed to navigate the car along the unknown track, a sensing system able to detect the sideway cones is needed as so path planning can be achieved.
To do so, an image processing algorithm, with or without the combination of a LiDAR sensor, able to detect fixed color and size cones and extract a distance and angle to each one (at least 4 per frame) at race-car speeds (as much as possible, no less then 20Kph) is required and is hence the goal of this project. The students will be an active part of the Formula Technion Driverless team (in good or bad).
3. Input –
Open to students’ discretion, either using one or more high-speed cameras and a LiDAR sensor and any other sensor thought to be fit.
4. Output –
Angle and distance vector for at least 4 upcoming cones (two at each side) at a frame rate high enough to allow the car to travel and speeds greater then 30Kph.
5. Algorithm base –
Open to students’ discretion, either classic image processing or with machine learning algorithms, as long as demands are met.
6. Duration–POC in early January and fully functioning system by march.
7. Number of students–Approx. 4 or as defined by lab director.
or: Johanan Erez firstname.lastname@example.org (VISL lab)