A nonlinear Stein based estimator for multichannel image denoising

The use of multicomponent images has become widespread with the improvement of multisensor systems having increased spatial and spectral resolutions. However, the observed images are often corrupted by an additive Gaussian noise. In this paper, we are interested in multichannel image denoising based on a multiscale representation of the images. A multivariate statistical approach is adopted to take into account both the spatial and the inter-component correlations existing between the different wavelet subbands. More precisely, we propose a new parametric nonlinear estimator which generalizes many reported denoising methods. The derivation of the optimal parameters is achieved by applying Stein’s principle in the multivariate case. Experiments performed on multispectral remote sensing images clearly indicate that our method outperforms conventional wavelet denoising techniques

💡 Research Summary

The paper addresses the problem of denoising multichannel (multispectral or hyperspectral) images that are corrupted by additive Gaussian noise. The authors adopt a multiscale wavelet representation, decomposing each image into sub‑bands that capture different scales and orientations. Unlike conventional approaches that treat each spectral band independently or apply simple inter‑band averaging, this work models the wavelet coefficients of all channels at a given sub‑band as a single multivariate random vector. Assuming a zero‑mean Gaussian noise with known variance, the observed coefficient vector (\mathbf{y}) is expressed as the sum of the true signal vector (\mathbf{x}) and noise (\mathbf{n}).

The central contribution is the extension of Stein’s unbiased risk estimator (SURE) to the multivariate case. By deriving a multivariate SURE formula, the authors obtain an unbiased estimate of the mean‑squared error (MSE) that depends only on the observed data. This enables the design of a parametric nonlinear estimator of the form
\