Algorithms For The Segmentation Of Speckled Images
- 25 August 2005
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 1 (10586393) , 279-284
- https://doi.org/10.1109/acssc.1988.753999
Abstract
This paper presents new algorithms for the segmentation of speckled images, including synthetic aperture radar (SAR) images. These algorithms are based on a hierarchical random field model proposed for speckled images and maximum a posteriori (MAP) segmentation using simulated annealing. Different versions of the model are developed to represent single- or multi-look and intensity or complex speckled images. In an earlier study, a similar segmentation algorithm was presented for single-look intensity speckled images. The algorithms in this paper treat single- and multi-look complex speckled images and multi-look intensity speckled images. These algorithms constitute an extension and an improvement over the earlier one. Parameters of the speckle processes are estimated as part of the algorithm. The segmentation algorithms are implemented on a wide class of synthetically generated and actual speckled images. The results verify that significant improvement is obtained when multi-look and complex image data is used for the segmentation.Keywords
This publication has 3 references indexed in Scilit:
- Adaptive segmentation of speckled images using a hierarchical random field modelIEEE Transactions on Acoustics, Speech, and Signal Processing, 1988
- Adaptive restoration of images with speckleIEEE Transactions on Acoustics, Speech, and Signal Processing, 1987
- Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of ImagesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1984