A syntactical approach to learn and identify bidimensional image models

Abstract

In one hand, automatic generation of models from a set of positive and negative samples and a-priori knowledge (if available) is a crucial issue for pattern recongnition applications. In the other hand, a generic multipurpose 2D object model representation is very useful in object recognition in complex scenes. In this paper we present a new approach of 2D objects multi-purpose model model representation based in context sensitive language and automatic learning. To illustrate the model representation of the performances achieved two different applications have been developed: an outdoor traffic sign identifier and a human face identifier. Partial results of the recognition process of both applications are shown.