SELF–ORGANIZING MAP COMBINED WITH A FUZZY CLUSTERING FOR COLOR IMAGE SEGMENTATION OF EDIBLE BEANS

Abstract

A novel segmentation approach that partitions color images into two uniform regions is described. Thisunsupervised procedure is based on a selforganizing map neural network and fuzzy cmeans clustering (SOM_FCM). Theselforganizing map allows the mapping of a color image related to edible beans into a consistent twodimensional tablethrough a nonlinear projection. Fuzzy clustering is then applied to the Kohonen map to determine the two cluster centers.The results were compared to a standard spatial thresholding segmentation method. The two segmentation approaches wereused for the segmentation of 150 color images of beans (acceptable, small, damaged, and broken), foreign materials, andstones. The results showed that the SOM_FCM outperformed the spatial thresholding method in identifying objects. It wasfound that the size of the Kohonen layer, the form of the neighborhood function, and the mapping topology did not have asignificant effect on the segmentation performance of the SOM_FCM. The average percentage of correctly matched pixelswas 99.31% for the SOM_FCM and only 89.71% for the spatial thresholding method. Unlike the SOM_FCM, the spatialthresholding method failed to correctly segment most of the broken bean and stone images. Unsupervised neural networkshave the potential to improve agricultural machine vision applications.