Melanoma Detection Using Decision Fusion of Various Classifiers

In this paper we present a simple and efficient approach based on the decision fusion of the characteristics-based method (ABCD) and deep neural networks-based method to classify skin lesions as melanoma or common nevus. As deep neural networks AlexNet and GoogLeNet were used. The decision fusion considers a convolutional rule of associated weights and probabilities provided by the previous classifiers. The method was implemented in the Matlab software and dermoscopic images from the PH2 and ISIC databases were used. The experimental results and the melanoma detection rate validate the proposed approach.