ing at problem that the single image clearness algorithm cannot guarantee the effect of noise and fog reduction, a new algorithm for foggy and noisy single image clearness using the lazy random walk model is proposed. Firstly, physical meaning of existing atmospheric scattering model is analyzed and improved to be made more consistent with the specialist of the actual foggy images. Lazy random walk model is used to estimate the attenuation term of improved haze degeneration model. Secondly, geometric constraint and color-line prior are used to obtain accurate atmospheric light of the degeneration model. Fog-free images with little noise are recovered at last. Experimental results show that the proposed algorithm can obtain the best defogging effect and restrain the noise, and it has strong robustness.

ing at the problem of feature redundancy in polarimetric synthetic aperture radar (SAR) application, a semi-supervised dimension reduction algorithm: semi-supervised local discriminant analysis (SLDA) is proposed by combining the thoughts of linear discriminant analysis (LDA) and locally linear embedding (LLE). Firstly, the regularization term is established based on local preserving property of LLE to avoid overfitting problem during learning. Then, discriminant analysis with regularization is performed on labeled data set in order to improve the generalization ability and preserve the local geometric structure in original space for the whole data. Dimension reduction experiments are performed on all polarimetric SAR data from Flevoland regions obtained by RADARSAT-2 and AIRSAR satellites. The results show that the low dimensional features extracted by SLDA has the characteristics of “intra compactness and inter separation”. Further classification experiment results show that SLDA can make the classification accuracy reach about 90% only with 1‰-2‰ labeled samples, and the classification performance of SLDA is superior to other comparison algorithms.