Fuzzy Based Morphological Shape Representation Scheme

Shape representation and description plays an important role in many areas of computer vision and pattern recognition. Mathematical morphology has emerged as a standard tool for detecting and isolating specific geometric structures in image data. Morphological shape decomposition techniques extract the structural components from the shape. The representation efficiency of the existing morphological shape decomposition algorithms is affected by two major factors such as the hard decision in decomposing the given shape in to simpler shape components and the effect of noise present in images. Fuzzy mathematical morphology overcomes these two problems by replacing the Boolean constructs in the morphological operations by corresponding generalizations from fuzzy logic. By app lying the powerful and flexible fuzzy concept for shape decomposition, the efficiency of morphological shape representation algorithms can be improved. We propose a new fuzzy based efficient morphological shape decomposition scheme (FMSD) which represents the object shape by a minimal set of shape components even with the presence of noise. 1. MORPHOLOGY BASED SHAPE REPRESENTATION APPROACHES One of the most important problems in image analysis is object representation and shape is the main feature used to represent the object. Mathematical morphology is the analysis of signals in terms of shape.[1] When applied to image processing, mathematical morphology has several advantages over other techniques outlined as preserves edge information, shape based processing and computationally efficient. Morphology based approaches examine the geometrical structure of an image by matching it with small patterns called structuring elements at various locations. By varying the size and shape of the matching patterns, useful structural components and their interrelations can be extracted which can be applied to image understanding, image indexing and machine vision.[2]. A binary shape X is represented as a union of shape components contained in X.