论文信息 - A new SAR image segmentation method based on fusion

A new SAR image segmentation method based on fusion

In this paper, a new synthetic aperture radar(SAR) image segmentation approach based on fusion is proposed. The wavelet transform algorithm and morphology reconstruction algorithm fusion technique can preserve more spatial feature and more functional information content, respectively. The new fusion method are adopted 2 SAR images of the same area taken with overexposure and under exposure are used for image processing. Compared with other methods, the experimental results of the proposed method based on image fusion using real SAR images are much better in reservation of information and boundary as well as correctness of original region.

Zhang Xin | Chen Wei-bin

[1] Richard G. Baraniuk,et al. Multiscale image segmentation using wavelet-domain hidden Markov models , 2001, IEEE Trans. Image Process..

[2] Reuven Meth. Target characterization and matching in synthetic aperture radar imagery , 1998 .

[3] Pierfrancesco Lombardo,et al. Optimum detection and segmentation of oil-slicks using polarimetric SAR data , 2000 .

[4] Pierfrancesco Lombardo,et al. Optimal classification of polarimetric SAR images using segmentation , 2002, Proceedings of the 2002 IEEE Radar Conference (IEEE Cat. No.02CH37322).

[5] Hong Sun,et al. An unsupervised segmentation method based on MPM for SAR images , 2005, IEEE Geoscience and Remote Sensing Letters.

[6] Michalis Zervakis,et al. A morphological fusion algorithm for optical detection and quantification of decay patterns on stone surfaces , 2008 .

[7] Rod Cook,et al. Optimal approach to SAR image segmentation and classification , 2000 .

[8] Guillermo Sapiro,et al. Knowledge-based segmentation of SAR data with learned priors , 2000, IEEE Trans. Image Process..