*PG Scholar ME, Avionics, PSN College of Engineering and Technology, Tirunelveli-152, India
**Associate Professor, Department of Aeronautical Engineering, PSN College of Engineering and Technology, Tirunelveli-152, India
Online published on 7 November, 2013.
The presence of speckle in Synthetic Aperture Radar (SAR) images makes the segmentation of such images difficult. A novel method for automatic segmentation of SAR images is proposed. Firstly, a wavelet packet based texture feature set is derived. It consists of the energy of the feature sub images obtained by the over complete wavelet packet decomposition of local areas in SAR image, where the down sampling between wavelet levels is omitted. Then an improved unsupervised clustering algorithm is proposed for image segmentation, which can determine the number of classes automatically. Segmentation results on real SAR image demonstrate the effectiveness of the proposed method.
Image segmentation, over complete wavelet packet decomposition, texture feature, clustering algorithm
