An ISAR Image Segmentaion Method Based on MMARP Model

This paper presents a method of unsupervised segmentation for Inverse synthetic aperture radar (ISAR) images. Firstly, a generalized multiresolution likelihood ratio (GMLR) is defined, which classifies different kinds of signals more accurately than classical likelihood ratio by fusing more and different signal features. For our ISAR image segmentation application, multiresolution stochastic structure inherent in ISAR imagery is well captured by a set of multiscale autoregressive (MAR) models. Secondly, good parameter estimates of GMLR can be obtained by estimating several mixture multiscale autoregressive prediction (MMARP) models using EM algorithm. Thirdly, considering the independence assumption of maximum likelihood estimation of parameter by EM algorithm and reduction of the segmentation time, we present the bootstrap sampling techniques applied above algorithm. Finally, Experimental results demonstrate that our algorithm performs fairly well.