Hours:
20 hours (5 credits)

Room:

Aula Riunioni del Dipartimento di Ingegneria dell’Informazione, Via G. Caruso 16, Pisa - Ground Floor

To register to the course, click here

Short Abstract:

AI and deep learning are improving and revolutionizing machine vision and, while 2D elaboration are almost all camera model agnostic, their 3D application require a deep understanding of the projective geometry and the basic concepts of traditional localization approaches.

This course offers an overview on the matter starting from the basic concepts up to some demos with the application of the most advanced AI powered localization algorithms. 

Course Contents in brief:

1. Camera modeling.
2. Tracking of points and structured markers with mono e stereo cameras.
3. 3D scan elements and their application to tracking.
4. Marker-less model (edge) based tracking.
5. Marker-less DL tracking: detection, instance and model aware approaches.

Schedule:

1. February 20, 2025 -  9:00-13:00  "Principles of geometric optics and camera model.Geometric and algebraic pinhole model of common cameras. Camera calibration algorithms and demo."
2. February 21, 2025 -  9:00-13:00  "PnP based tracking of structured planar markers.Tracking of planar unstructured markers. Tracking accuracy of planar markers."
3. February 25, 2025 -  14:00-18:00  "Multiview tracking.Accuracy of stereo tracking. 3D scan elements and their application to tracking."
4. February 27, 2025 -  14:00-18:00  "Introduction to marker less tracking.Model (edge) based tracking. Marker-less DL tracking: detection, instance and model aware approaches."
5. February 28, 2025 -  9:00-13:00  "Exam".