Fuzzy-probabilistic object recognition based on edge and texture descriptors fusion

A great deal of research effort has been devoted to information processing in pattern recognition. It has been established that when it comes to making decisions upon the perceptual content of the information there is no case-independent and commonly approved technique. This is in part due to the complexity of the real world information to the measurements´ ambiguity or the observer´s imperfect knowledge, but mostly, to the non-integrity of the available information. Fusing many knowledge sources is a recent trend in the pattern recognition field. Fuzzy logic provides a general conceptual framework and an analytical tool as well. Fusion and fuzzy uncertain modelling of salient image features such as edges and texture are at the core of the proposed method