Automatic Target Recognition Using Multiple Description Coding Models for Multiple Classifier Systems

Abstract

In this paper, we proposed a new multiple classifier system (MCS) based on multiple description coding (MDC) models. Our proposed method was inspired from the framework of transmitting data over heterogeneous network, especially wireless network. In order to support the idea of MDC in pattern classification, parallels between transmission of concepts (hypothesis) and transmission of information through a noisy channel are addressed. Furthermore, preliminary surveys on the biological plausible of the MDC concepts are also included. One of the benefits of our approach is that it allows us to formulate a generalized class of signal processing based weak classification algorithms. This will be very applicable for MCS in high dimensional classification problems, such as image recognition. Performance results for automatic target recognition are presented for synthetic aperture radar (SAR) images from the MSTAR public release data set.