Academic journal article Issues in Informing Science & Information Technology

Modified Watershed Algorithm for Segmentation of 2D Images

Academic journal article Issues in Informing Science & Information Technology

Modified Watershed Algorithm for Segmentation of 2D Images

Article excerpt


In image processing, segmentation is a basic problem in different fields for example, pattern recognition, scene analysis and image analysis. Image segmentation is the process of dividing images into regions according to its characteristic e.g., color and objects present in the images. These regions are sets of pixels and have some meaningful information about object. The result of image segmentation is in the form of images that are more meaningful, easier to understand and easier to analyze. In order to locate objects and boundaries in images feature extraction of object shape, optical density, and texture, surface visualization, image registration and compression image segmentation is used. Correct segmented results are very useful for the analysis, predication and diagnoses (Segmentation (image processing), n.d.).

Watershed segmentation is a morphological based method of image segmentation. The gradient magnitude of an image is considered as a topographic surface for the watershed transformation. Watershed lines can be found by different ways. The complete division of the image through watershed transformation relies mostly on a good estimation of image gradients. The result of the watershed transform is degraded by the background noise and produces the over-segmentation. Also, under segmentation is produced by low-contrast edges generate small magnitude gradients, causing distinct regions to be erroneously merged.

In order to reduce the deficiencies of watershed, many preprocessing techniques are proposed by the different researchers for example Jung and Scharcanski (2005) presents a robust watershed segmentation using wavelets where wavelets technique is used to denoise the image. Bieniek and Moga (2000) present an efficient watershed algorithm based on connected components. Hamarneh and Li (2009) have proposed a method of watershed segmentation using prior shape and appearance knowledge to improve the segmentation results etc. But most of the techniques previously proposed consider the over segmentation problems and focus on the denoising of the image. The image low contrast and under segmentation problem is not yet addressed by most of the researchers.

The proposed technique focuses on the solution of under segmentation problem of low contrast images by applying preprocessing on the input image. The technique for preprocessing on the images is Random Walk (Erikson, 2005). It is probabilistic approach used to enhance the image contrast in the way image is degraded.

The division of this paper is as follows, in section 2 some related work is given which describes the previous research about the remedy of watershed issues. In section 3, some basic information about watershed and random walk is given. In section 4, the proposed technique is discussed. In section 5, results are given. In section 6, the conclusion and future plane is given.

Related Work

Erikson (2005) proposes two different techniques of performing preprocessing of an image to improve segmentation results. The methods use the grey level thickness of the objects, in order to find the resulting image. The first method proposed by the author is RW, uses the random walk of a particle to a random position, the position is defined in the neighborhood of the particle. The resulting image through this method holds the number of times the particle visits a pixel. The second method is Iterative Procedure (IP) scans the image iteratively and calculates the expected value of the same number, instead of randomization to find the number of visits. The methods proposed in this paper are independent of the segmentation method and can therefore be used as a preprocessing step for other segmentation methods as well.

Beucher (1991) proposed a method for image segmentation based on the mathematical morphology. The process of image segmentation is divides into two approaches, boundary based and region based. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.