SAR Image Target Recognition with Increasing Sub-classifier Diversity Based on Adaptive Boosting

Adaptive boosting (AdaBoost) can boost a weak learning algorithm with an accuracy slightly better than random guessing into an arbitrarily accurate strong learning algorithm. It has been applied to target recognition with single feature classifier. However, the poor discriminative power of extremely weak single feature classifier limits its application. In this paper, we present a novel comprehensive classifier design method toward synthetic aperture radar (SAR) image target recognition with increasing sub classifier diversity based on AdaBoost, we call it AdaBoost. ISCD. First, it trains weak classifiers with support vector machine (SVM) and k-Nearest Neighbor (k-NN) by extracting different features of the training samples. Then, it constitutes a strong classifier by combining weak classifiers with AdaBoost. Experimental results based on SAR images show that the presented method is a feasible and effective recognition method in terms of classification accuracy.