Denoising is an important preprocessing step to further analyze a hyperspectral image (HSI). The common denoising methods, such as 2-D (two dimensions) filter, actually rearrange data from 3-D to 2-D and ignore the spectral relationships among different bands of the image. Tensor decomposition method has been adapted to denoise HSIs, for instance Tucker3 (three-mode factor analysis) model. However, this model has problems in uniqueness of decomposition and in estimation of multiple ranks. In this paper, to overcome these problems, we exploit a powerful multilinear algebra model, named parallel factor analysis (PARAFAC), and the number of estimated rank is reduced to one. Assumed that HSI is disturbed by white Gaussian noise, the optimal rank of PARAFAC is estimated according to that the covariance matrix of the n-mode unfolding matrix of the removed noise should be approach to a scalar matrix. Then, the denoising results by PARAFAC decomposition are presented and compared with those obtained by Tucker3 model and 2-D filters. To further verify the denoising performance of PARAFAC decomposition, Cramer-Rao lower bound (CRLB) of denoising is deduced theoretically for the first time, and the experiment results show that the PARAFAC model is a preferable denoising method since the variance of the HSI denoised by it is closer to the CRLB than by other considered methods.Index Terms-Classification, Cramer-Rao lower bound (CRLB), denoising, hyperspectral image (HSI), parallel factor analysis (PARAFAC), Tucker3.

SCD patients in Cameroon presented a very high prevalence of cognitive deficits, with a specific impairment of executive functions and attention. Routine neuropsychological evaluation for early detection of cognitive deficits in SCD patients could represent a cost-effective tool to implement in resource-limited contexts such as in sub-Saharan Africa.

Hand posture recognition remains a challenging task for in-line systems working directly in the video stream. In this work, we compare several shape descriptors, with the objective of finding a good compromise between accuracy of recognition and computation load for a real-time application. Experiments are run on two families of contour-based Fourier descriptors and two sets of region-based moments, all of them are invariant to translation, rotation and scale changes of hands. These methods are independent of the camera view point. Systematic tests are performed on the Triesch benchmark database and on our own large database, which includes more realistic conditions. Temporal filtering and a method for unknown posture detection are considered to improve posture recognition results in case of video stream processing.

The main goal of the method proposed in this paper is the numerical study of various kinds of anisotropic gratings deposited on isotropic substrates, without any constraint upon the diffractive pattern geometry or electromagnetic properties. To that end we propose a new FEM (Finite Element Method) formulation which rigorously deals with each infinite issue inherent to grating problems. As an example, 2D numerical experiments are presented in the cases of the diffraction of a plane wave by an anisotropic aragonite grating on silica substrate (for the two polarization cases and at normal or oblique incidence). We emphasize the interesting property that the diffracted field is non symmetric in a geometrically symmetric configuration.

No abstract