Site icon G. Mancusi's blog

Paintings recognition, People detection in a Museum.

This work lets you recognize paintings in a museum from photos and videos, allows you to rectify them when they are seen in perspective, allows you to get information about the original painting. It is also integrated a people detection system and identification of the room in which they are located. We have experimented with the use of neural networks and at the same time, we have used classic algorithms.

The project was designed to complete the university exam of Vision and Cognitive Systems for professors Rita Cucchiara and Lorenzo Baraldi. The team is composed of two other people, Daniele Manicardi and Vittorio Pippi.

Technical details can be found on this PDF:

The project and source code can be found at this link on GitHub:


In this image you can see how the camera was calibrated automatically, we recognized the paintings and the room they are in.
The original image.

Exit mobile version