A fuzzy based texture retrieval system that considers psychological aspects

Abstract

Texture perception plays an important role in human vision. It is used to detect and distinguish objects, to infer surface orientation and perspective, and to determine shape in 3D scenes. An interesting psychological observation is the fact that humans' beings are not able to describe textures clearly and objectively, but only subjectively by using a fuzzy characterisation of them. On the other hand, with the new advances in communication and multimedia computing technologies, accessing mass amounts of digital visual information (image databases) is becoming a reality. In this context, textures, due to their aesthetical properties, play today an important role in the consumer-oriented design, marketing, selling and exchange of products and/or product information. For this reason, systems that allow the search and retrieval of textures in image databases, the so called Texture Retrieval Systems, are of increasing interest. The propose is to describe a new Texture Retrieval System 1, which is based on the use of Fuzzy Logic, Neuro-Fuzzy Networks and Morphological Operators in the processes of Qualitative to Quantitative Textural Properties Transformation, Color and Textural Features Extraction, Features Fusion and Feature Similarity Matching. One important aspect of the proposed system is that it considers psychological aspects of description and perception of textures. A textural retrieval process can be divided into two main steps: a) off-line generation of the image annotations in the database, and b) on-line retrieval of textured images from the database. The proposed retrieval system, whose block diagram is shown in Fig. 1, is made of the QZTPT (Qualitative to Quantitative Textural Properties Transformation), the FSM (Feature Similarity Matching) and the FTD (Features/Texture Database) modules. 1 Demo program under the following electronic address: http:/Istrauss.ipk.fhg.de/Textursynthese/texql.html