MMW Image Blind Restoration Using Sparse ICA in Contourlet Transform Domain
Sparse independent component analysis (SPICA) algorithm is effective in blind separation of superimposed images, without having any priory knowledge about the image’s structure and statistics. While a millimeter wave (MMW) image contains the refective information of imaging object and much unknown noise of imaging scene, so the MMW image is too high blur to be discerned. To obtain preferable MMW image, combined the advantages of contourlet sparse transform and SPICA, a new blind restoration method proposed by us of MMW images operating in the contourlet sparse transform domain is discussed in this paper. Contourlet transform can retain the better contour of an image and make this image sparser in local subspace. Here, using the low frequency band and the high frequency bands of the first layer obtained by contourlet transform as the mixed input data of SPICA, the task of MMW image restoration can be implemented. In test, the blind restoration of mixed natural images is also operated by using our method, simultaneity, using the single noise ratio (SNR) to measure the restored natural images, experimental results testify the validity of our method in doing blind separation and it is feasible to restore the MMW image using this proposed method. Further, compared with methods of contourlet transform and fast ICA, simulations again show that this MMW image restoration method proposed is indeed efficient in application.
KeywordsMillimeter wave (MMW) Independent component analysis (ICA) Sparse ICA Contourlet sparse transform Image restoration
Unable to display preview. Download preview PDF.
- 2.Sundareshan, M.K.: Bhattacharjee Supratik: Superresolution of Passive Millimeter-wave Images Using a Combined Maximum-likelihood Optimization and Projection-onto-convex-sets Approach. In: Proc. of SPIE Conf. on Passive Millimeter-wave Imaging Technology, Acrosense 2001, Orlando, FL, UAS, vol. 4373, pp. 105–116 (2001)Google Scholar
- 3.Cheng, P., Zhao, J.Q., Si, X.C., et al.: L-R Imaging Algorithm For Passive Millimeter Wave Based on Sparse Representation. Journal of Electronics & Information Technology 32, 1707–1711 (2010)Google Scholar
- 4.Hyvärinen, A., Karhunen, J.H., Oja, E., et al.: Independent Component Analysis. A Wiley-Interscience Publication, New York (1999)Google Scholar
- 5.Zhang, K., Chan, L.W.: ICA with Sparse Connections. Intelligent Data Engineering and Automated Learning 9, 530–537 (2006)Google Scholar