A novel method for image segmentation using pulse-coupled neural network based on root mean square of gray scale

Abstract

PCNN (pulse-coupled neural network) has good properties for image segmentation. A novel method which integrates root mean square of grayscale threshold with PCNN, called root mean square of grayscale threshold PCNN (RMS-PCNN), is proposed. In simplified method there is only one parameter to be determined. Furthermore, the PCNN threshold is computed once so that the image is segmented during an ignition process. Better performance and faster computation are demonstrated. Experimental results validate the effectively simplified method on segmentation.