Using a Computational Model of Human Color Vision to Perform Object Segmentation in Natural Images