This paper addresses the issue of image segmentation. Image segmentation process is the main basic process or technique used in various image processing problem domains, for example, content based image retrieval; pattern recognition; object recognition; face recognition; medical image processing; fault detection in product industries; etc. Scope of improvement exists in the following areas: Image partitioning; color based feature; texture based feature, searching mechanism for similarity; cluster formation logic; pixel connectivity criterion; intelligent decision making for clustering; processing time; etc. This paper presents the image segmentation mechanism which addresses few of the identified areas where the scope of contribution exists. Presented work basically deals with the development of the mechanism which is divided into three parts first part focuses on the color based image segmentation using k-means clustering methodology. Second part deals with region properties based segmentation. Later, deals with the boundary based segmentation. In all these three approaches, finally the Steiner tree is created to identify the class of the region. For this purpose the Euclidean distance is used. Experimental result justifies the application of the developed mechanism for the image segmentation.
Ranit, Snehal B. and THAKUR, NILESHSINGH V.
"Image Segmentation using Various Approaches,"
International Journal of Image Processing and Vision Science: Vol. 2:
2, Article 6.
Available at: https://www.interscience.in/ijipvs/vol2/iss2/6