Citation Context

...ly, Maji et al. [16] proposed the biased normalized cut, a modification of the normalized cut to incorporate priors which can be used for constrained color-texture based image segmentation. In this paper a genetic algorithm for color image segmentation that adopts the representation of an image as a graph is proposed. The algorithm, named C-GeNCut (Color A.I. Esparcia-Alcazar et al. (Eds.): EvoApplications 2013, LNCS 7835, pp. 314–323, 2013. c© Springer-Verlag Berlin Heidelberg 2013 A Genetic Algorithm for Color Image Segmentation 315 Genetic Normalized Cut), extends the approach proposed in [1] to take into account color, brightness, and texture, by computing the affinity between two pixels with the Intervening Contour cue [8, 10, 11, 15], that uses the multispectral Pb detector as defined by Arbelaez et al. in [2]. Experiments on ten color images from the Berkeley Image Segmentation Dataset (BSDS300) [17] show that the method is able to partition natural and human scenes in meaningful objects. A quantitative evaluation based on the well known concept of Probabilistic Rand Index, defined by Unnikrishnan et al. [21, 22], shows that the inclusion of color and texture improves the segm...

Citation Context

...s have been developed based onsclustering as in [19],[21],[22],[23], threshold based fuzzyslogic as in [20], neural network as in [24],[26], geneticsalgorithm as in [1],[10], weighted graph as in [2],=-=[25]-=-,sevolutionary computation as in [18] etc.sIn this paper, a new approach to perform targetedssegmentation has been proposed which will be using ascombination of decision making and neuro-fuzzystechniq...