期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2014
卷号：69
期号：3
出版社：Journal of Theoretical and Applied
摘要：Image segmentation techniques are discussed earlier which uses color, intensity, pixel values and etc, and suffers with the problem of overlapping and grouping efficiency. We propose a novel method for image segmentation, using multi variant feature based energy estimation technique and works with interactive mode. Unlike other methods, the proposed methodology computes the feature energy of each pixel. The Feature energy represents the propagation of feature from one pixel to another and energy transition occurred at each of the neighbor pixel. The energy estimation is performed using various features of each pixel and its neighbors. An each pixel becomes the member of a group only if it carries same set of energy level or same set of features with possible little changes and comes within the energy threshold. The features used to compute the energy level are pixel values like red, green, blue and the number of neighbors and their values. The computed value is used to perform segmentation process, and produces higher rate of segmentation. The process will be performed iteratively still the user gets satisfied and provides new energy threshold value.
