THESAURUS - Report D3.3: Metodi per il riconoscimento e l'annotazione automatica di immagini e mappe

This report contains a description of the software design and development activity in the framework of the Thesaurus project experimental tests. The marine environment represents a challenging context for IT experts, both for what is concerning the discovery of unknown sites and for the recovery and preservation operations. The hostile environmental conditions, unfit to human intervention, favored the increasing demand for artificial intelligence integration in the AUVs, the Autonomous Underwater Vehicles in charge of survey operations. The primary purpose of our work consists in treating optical and acoustic signals in order to extract useful information about the surveyed site. We basically seek for meaningful features in the data in order to perform robust object recognition, also in case of unfavorable environmental conditions (in presence of noise, partial occlusions, etc...). The report is organized as follows: the project experimental conditions are described in section 2; in section 3 the techniques employed for a preliminary signal processing are described, paying specific attention to the difference between optical and acoustic data; the procedures performing automatic recognition and classification are described in the following 4,5 and 6 sections; finally the implemented data fusion model is described in section 7. The report contains also some of the results obtained by processing simulation data as well as the data captured during the project experimental tests.