- Daniel Paez Martin (212600)
This project targets an autonomous driving scenario under an intelligent autonomous systems perspective. The scenario considers a dynamic environment, where vehicles, pedestrians and other typical urban entities can coexist without risk. Such a dynamic and unstructured environment can result in occlusions in the sensors of the vision-based autonomous driving systems. For this project, detection and tracking of entities of interest is performed with a stereo camera that is fixed on the car, looking forward. Objects in the environment in front of the car need to be tracked in 3D, even if occlusions happen, and be classified in their respective object type.
| Topics | Name |
|---|---|
| Image Processing, Feature Detection, Description, and Matching, Optical Flow | Daniel |
| Multiple View Geometry: Disparity, Calibration, Fundamental Matrix Essential Matrix, 3D * Reconstruction | Denisa |
| 3D Point Cloud Processing: Registration, Regression, Clustering | Subramanya |
| State Estimation: Kalman Filter | Tianyu |
| Classification | Navaneeth |
| Visual Odometry, Motion Estimation | Medhavi |
