Recognition and classification of geometrical objects using bio-inspired vision techniques
General description (stub)
The goal of the project is to investigate the recognition of geometrical objects by the human vision system to replicate its functional model on a robot. Many researches in the neurophysiology field show that some simple geometrical objects are automatically recognized by the neural network without the help of the cognitive system.
The secondary goal is to build an autonomous ontology-based classification system of the recognized objects; every recognized object should be assigned to a symbol, making the robot able to perform symbolic computation. The obtained symbolic information can be successively matched with an ad-hoc ontology representing physical properties of some materials.