INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS www.ijrcar.com Vol.4 Issue 1, Pg.: 38-55 January 2016 D. Chaudhuri Page 38 INTERNATIONAL JOURNAL OF RESEARCH IN COMPUTER APPLICATIONS AND ROBOTICS ISSN 2320-7345 IMAGE ANALYSIS USING A NEW DEFINITION OF MATHEMATICAL MORPHOLOGY FOR BINARY IMAGE D. Chaudhuri DRDO Integration Centre, Panagarh, Burdwan West Bengal, 713419, INDIA E-mail: deba_chaudhuri@yahoo.co.in Abstract: - What the algebra of convolution does for linear systems, the algebra of mathematical morphology does for shape. Since shape is a prime carrier of information in machine vision, there should be little surprise about the importance of the mathematical morphology. Now adays the operations of mathematical morphology are very much useful in object or defect identification required in industrial vision applications and target detection useful to military applications. A new definition for binary morphology covering the operation of dilation, erosion, opening and closing and their relations are presented in this paper. Examples are presented for each morphological operation. Keywords: Morphology, dilation, erosion, opening, closing, shape analysis. 1. Introduction Mathematical morphology has evolved as a useful tool for various image-processing tasks (Giardina & Daugherty, 1988), (Serra, 1982). An algebraic system of operators, such as those of mathematical morphology, is useful to the processing of digital images that are based on shape. “Morphology” can literally be taken to mean, “doing things to shapes”. “Mathematical morphology” then, by extension, means using mathematical principals to do things to shapes. It treats an image as an ensemble of sets rather than signal. Its language is that of the set theory and operations are defined in terms of the iterations between the object and the structuring elements. The operators are able to decompose a complex shape structure into its meaningful parts and separate the meaningful parts from its extraneous parts. Morphological operators and their compositions are able to identify the underline shapes and reconstruct the best possible from their distorted noisy forms. Morphological operations can simplify image data, preserving their essential shape characteristics and eliminates irrelevancies. As the identification and decomposition of objects, object features, object surface defects and assembly defects correlate directly with shape; it is only natural that mathematical morphology has an essential structural role to play in machine vision (Haralick ei al., 1987), (Serra, 1986). Majman and Tolbot (2010) have presented a survey of the state of the art in mathematicl morphology. Now a day mathematical morphological tools are one of the most effective tool in computer vision and morphological image processing is currently an emerging research topic (Tankyerych et al., 2013), (Miranda and Mansilla, 2014), (Ronse, 2014). Nakagawa and Rosenfeld (1978) first discussed the use of neighbourhood minimum and maximum operators for shrinking and expanding operations on two-valued digital pictures, which are useful for noise removal, as well as for detecting dense regions and elongated parts of objects. Sternberg (1982) has extended the work in generalised form. Peleg and Rosenfeld (1981) use greyscale morphology to generalise the medial axis transform to greyscale imaging. Tolbot and Appleton (2007) have presented a new ordered implementations of the complete and incomplete path opening and closing operators. Peleg et al. (1984) use greyscale morphology to measure changes in texture properties as a function of resolution. Werman and Peleg (1985) use greyscale morphology for texture feature extraction. Favre at al. (1985) use greyscale