A Computer Vision System for Monitoring Ice-Creams Freezers

Abstract

In this thesis, it will be presented a set of computer vision algorithms mainly devoted to monitor the evolution of the content inside an ice-creams freezer for commercial purposes. In our work we have been asked to address tasks such as ice-creams baskets detection, ice-creams volume checking, products tracking and products matching, with different required degrees of accuracy. The information about the freezer was provided by four fish-eye cameras placed inside and connected to an embedded mini-computer able to transmit data by means of a telemetry system. We obtained promising results for most of the tasks in our laboratory analysis.