Hand skin color segmentation using online training artificial neural networks

A new approach for the hand image segmentation is presented in this paper. This method integrates the merits of histogram threshold, seeded region growing (SRG) and the Restricted Coulomb Energy (RCE) neural network. The performance of our proposed method is quantitatively compared with these three methods conducted independently. Experimental results have shown that our method is effective for the segmentation, which can be further used in the gesture-based human-computer interaction.