Ricerca CERCA

Who is what


Description The goal of the thesis is to identify a series of characteristics, such as the color of the uniform, skin, shoes and jersey number, in order to uniquely identify the athletes. These characteristics will be used as parameters to build a profile of the individual player. The algorithm that recognizes these characteristics, and consequently the individual athlete, needs to be developed as well.
If a sufficiently stable algorithm is developed, one of the long-standing object recognition issue would be solved: it will be able to tell who is what. Nowadays a neural network is only able to recognize all the players in a football field but cannot identify each individual player.
Thesis will be developed in collaboration with the company netventure r&d S.r.l., Ing. Antonio Bleile .
Basic skills recommended, or skills that will be improved through the thesis: C/C++, Computer Vision.

Deadline 23/05/2023      PROPONI LA TUA CANDIDATURA

© Politecnico di Torino
Corso Duca degli Abruzzi, 24 - 10129 Torino, ITALY