A practical approach to continuos-tone color image segmentation is proposed.Unlike traditional algorithms of image segmentation which tend to use threshold methods we intend to show how neural network technique can be successfully applied to this problem.We used a Bacpropagation network architecture in this work. It was assumed that each image pixel has its own color,which is somehow correlated with those of the nearest neighborhood. To describe the color properties of certain neighborhood we suggested nine component feature vector for every image pixel.This set offeature components is applied to the network input neurons. By this means,every image pixel is described by the following values R,G and B (color intensities), Mr, Mg and Mb (averages ofintensities of the nearest neighborhood), r ,Tgland b (.r. .m. s. deviations of color intensities). To estimate the algorithm efficiency the scalar criterion was proposed. It was shown by the results of comparative experiment that neural segmentation provides more efficiency then that oftraditional , using threshold methods.

No abstract