PROJECT TITLE :

Classification on the Monogenic Scale Space: Application to Target Recognition in SAR Image - 2015

ABSTRACT:

This paper introduces a unique classification strategy based on the monogenic scale space for target recognition in Artificial Aperture Radar (SAR) image. The proposed methodology exploits monogenic signal theory, a multidimensional generalization of the analytic signal, to capture the characteristics of SAR image, e.g., broad spectral info and simultaneous spatial localization. The components derived from the monogenic signal at totally different scales are then applied into a recently developed framework, sparse illustration-primarily based classification (SRC). Moreover, to house the information set, whose target classes are not linearly separable, the classification via kernel combination is proposed, where the multiple elements of the monogenic signal are jointly thought-about into a unifying framework for target recognition. The novelty of this paper comes from: the event of monogenic feature via uniformly downsampling, normalization, and concatenation of the elements at varied scales; the development of score-level fusion for SRCs; and the event of composite kernel learning for classification. In explicit, the comparative experimental studies underneath nonliteral operating conditions, e.g., structural modifications, random noise corruption, and variations in depression angle, are performed. The comparative experimental studies of numerous algorithms, as well as the linear support vector machine and therefore the kernel version, the SRC and also the variants, kernel SRC, kernel linear illustration, and sparse representation of monogenic signal, are performed too. The feasibility of the proposed methodology has been successfully verified using Moving and Stationary Target Acquiration and Recognition database. The experimental results demonstrate that significant improvement for recognition accuracy will be achieved by the proposed method compared with the baseline algorithms.