ESKMod, a CommonKADS Knowledge Model Integrating Multiple Classic Edge Based Segmentation Algorithms

Bibliography offers us a wide set of image segmentation methods, which are usually oriented to solve specific problems. These are evaluated, mainly, in specific work fields, as the analysis of industrial pieces, classification of agricultural products or remote sensing. Actually, main efforts are concentrated in the definition of new algorithms, generating a wider collection of alternative methods. Taking account of our experience about applying segmentation methods to diverse work fields, we realized the advantages of integrating previously used methods into a single segmentation model, with multiple alternatives. In this work, we present how some classic methods are integrated into a single segmentation knowledge model, using the CommonKADS modeling tools. Our final objective is building a model of the applicatibility of the alternatives in order to relate them with searched results. Here, we present how this composed knowledge model was assembled.