VISUAL OBJECT RECOGNITION

The aim if this project was to design a model that could recognise an object independently of its representation. An object representation may vary by applying different transformations to the object, such as changes of size, position, and rotation. A new approach to solve these problems has recently arisen from the constructions of Neural Network based on the anatomical and physiological properties of cortex. In this work, a survey of biological properties for vision will be made, as well as review of existing models that reproduce those properties. Furthermore, a new family of models will be presented and implemented with a corresponding measure of the properties