BibTeX:

@article{IJIRSTV2I5059,
     title={Unsupervised Change Detection in SAR Image based on Contourlet Fusion and Guided Filter },
     author={Priyamol James and Sinu Maria Kurian},
     journal={International Journal for Innovative Research in Science & Technology},
     volume={2},
     number={5},
     pages={196--203},
     year={},
     url={http://www.ijirst.org/articles/IJIRSTV2I5059.pdf},
     publisher={IJIRST (International Journal for Innovative Research in Science & Technology)},
}

A novel approach to unsupervised change detection method in synthetic aperture radar (SAR) images based on image fusion and compressed projection. Different types of terrain changes can be identified by using SAR images. First, proposed approach to generate a difference image by using Gauss-log ratio. The Gauss-log ratio operator is proposed to obtain a stable and clear difference map. In order to make better use of the Gauss-log ratio operator and log-ratio operator, this paper introduces a fusion strategy to generate a fused difference image based on Contour let Transform and Guided filter. Second non subsampled contour let transform (NSCT) to reduce the noise in the fused difference image and compressed projection is employed to extract feature for each pixel, improve the performance of the subsequent clustering or classification, NSCT model is used to suppress noise of the image and keep shape of the changed portion. The final change detection map is obtained by partitioning the feature vectors into “changed” and “unchanged” classes using simple k-means clustering. At last the experiment shows the performance evaluation of the different fusion strategy, using the PSNR value.

Discrete Wavelet Transform (DWT), Ratio Operators, Non subsampled contour let transform (NSCT), Guided Filter, K-mean clustering