Hemant Kumar Aggarwal

Hyperspectral Image Denoising Using Spatio-Spectral Total Variation

This letter introduces a hyperspectral denoising algorithm based on spatio-spectral total variation. The denoising problem has been formulated as a mixed noise reduction problem. A general noise model has been considered which accounts for not only Gaussian noise but also sparse noise. The inherent structure of hyperspectral images has been exploited by utilizing 2-D total variation along the spatial dimension and 1-D total variation along the spectral dimension. The denoising problem has been formulated as an optimization problem whose solution has been derived using the split-Bregman approach. Experimental results demonstrate that the proposed algorithm is able to reduce a significant amount of noise from real noisy hyperspectral images. The proposed algorithm has been compared with existing state-of-the-art approaches. The quantitative and qualitative results demonstrate the superiority of the proposed algorithm in terms of peak signal-to-noise ratio, structural similarity, and the visual quality.

Hyperspectral Unmixing in the Presence of Mixed Noise Using Joint-Sparsity and Total Variation

Hyperspectral unmixing is the process of estimating constituent endmembers and their fractional abundances present at each pixel in a hyperspectral image. A hyperspectral image is often corrupted by several kinds of noise. This work addresses the hyperspectral unmixing problem in a general scenario that considers the presence of mixed noise. The unmixing model explicitly takes into account both Gaussian noise and sparse noise. The unmixing problem has been formulated to exploit joint-sparsity of abundance maps. A total-variation-based regularization has also been utilized for modeling smoothness of abundance maps. The split-Bregman technique has been utilized to derive an algorithm for solving resulting optimization problem. Detailed experimental results on both synthetic and real hyperspectral images demonstrate the advantages of proposed technique.

Impulse denoising for hyper-spectral images

In this work we propose a technique to remove sparse impulse noise from hyperspectral images. Our algorithm accounts for the spatial redundancy and spectral correlation of such images. The proposed method is based on the recently introduced Blind Compressed Sensing (BCS) framework, i.e. it empirically learns the spatial and spectral sparsifying dictionaries while denoising the images. The BCS framework differs from existing CS techniques that employ fixed sparsifying basis; BCS also differs from prior dictionary learning studies which learn the dictionary in an offline training phase. Our proposed formulation has shown over 5dB improvement in PSNR over other techniques.

Exploiting spatiospectral correlation for impulse denoising in hyperspectral images

Abstract. This paper proposes a technique for reducing impulse noise from corrupted hyperspectral images. We exploit the spatiospectral correlation present in hyperspectral images to sparsify the datacube. Since impulse noise is sparse, denoising is framed as an ℓ1-norm regularized ℓ1-norm data fidelity minimization problem. We derive an efficient split Bregman-based algorithm to solve the same. Experiments on real datasets show that our proposed technique, when compared with state-of-the-art denoising algorithms, yields better results.

Single-sensor multi-spectral image demosaicing algorithm using learned interpolation weights

Multi-spectral images capture more information about a scene as compared to RGB images and have various scientific applications. But the high resolution multi-spectral cameras are very expensive which limits their wide applicability as compared to normal digital RGB cameras. In this paper a multi-spectral filter array design is proposed to capture multiple bands using the single-sensor architecture. The use of single-sensor can help in reducing the cost and size of multi-spectral cameras while simultaneously eliminating the image registration problem. Fast linear demosaicing technique is also proposed to interpolate missing values from under-sampled raw image. Experimental results show the superiority of proposed technique over state of art multi-spectral demosaicing technique.

Compressive sensing multi-spectral demosaicing from single sensor architecture

This paper addresses the recovery of multi-spectral images from single sensor cameras using compressed sensing (CS) techniques. It is an exploratory work since this particular problem has not been addressed before. We considered two types of sensor arrays - uniform and random; and two recovery approaches - Kronecker CS (KCS) and group-sparse reconstruction. Two sets of experiments were carried out. From the first set of experiments we find that both KCS and group-sparse recovery yields good results for random sampling, but for uniform sampling only KCS yields good results. In the second set of experiments we compared our proposed techniques with state-of-the-art methods. We find that our proposed methods yields considerable better results.

Removing sparse noise from hyperspectral images with sparse and low-rank penalties

In diffraction grating, at times, there are defective pixels on the focal plane array; this results in horizontal lines of corrupted pixels in some channels. Since only a few such pixels exist, the corruption/noise is sparse. Studies on sparse noise removal from hyperspectral noise are parsimonious. To remove such sparse noise, a prior work exploited the interband spectral correlation along with intraband spatial redundancy to yield a sparse representation in transform domains. We improve upon the prior technique. The intraband spatial redundancy is modeled as a sparse set of transform coefficients and the interband spectral correlation is modeled as a rank deficient matrix. The resulting optimization problem is solved using the split Bregman technique. Comparative experimental results show that our proposed approach is better than the previous one.

Mixed Gaussian and impulse denoising of hyperspectral images

Hyperspectral image denoising is an important preprocessing step in the analysis of hyperspectral images in several applicaitons domains. These images often gets corrupted by various kinds of noise during acquisition process. There are several studies on reducing Gaussian noise from hyperspectral images. This work addresses the problem of reducing mixed noise from hyperspectral images; in particular a mixture of Gaussian and impulse noise has been considered. The proposed image acquisition model explicitly accounts for both Gaussian and impulse noise as additive noise. This mixed noise reduction problem has been formulated as synthesis prior optimization problem which exploits inherent spatio-spectral correlation present in hyperspectral images. Split-Bregman based approach has been utilized to solve resulting optimization problem. Experiements were conducted using both synthetic noise as well as real noisy hyperspectral images. Experimental results have been quantified using peak signal to noise ratio (PSNR) and structural similarity index (SSIM). A comparative study with an existing low-rank based image denoising approaches has also been carried out. Both quantitative and qualitative results suggest the superiority of proposed approach.

Discriminative Robust Deep Dictionary Learning for Hyperspectral Image Classification

This paper proposes a new framework for deep learning that has been particularly tailored for hyperspectral image classification. We learn multiple levels of dictionaries in a robust fashion. The last layer is discriminative that learns a linear classifier. The training proceeds greedily; at a time, a single level of dictionary is learned and the coefficients used to train the next level. The coefficients from the final level are used for classification. Robustness is incorporated by minimizing the absolute deviations instead of the more popular Euclidean norm. The inbuilt robustness helps combat mixed noise (Gaussian and sparse) present in hyperspectral images. Results show that our proposed techniques outperform all other deep learning methods—deep belief network, stacked autoencoder, and convolutional neural network. The experiments have been carried out on both benchmark deep learning data sets (MNIST, CIFAR-10, and Street View House Numbers) as well as on real hyperspectral imaging data sets.

Generalized Synthesis and Analysis Prior Algorithms with Application to Impulse Denoising

This work proposes generalized synthesis and analysis prior algorithms using the split-Bregman technique for applications in impulse noise reduction. Impulse denoising is formulated as minimizing a Lp-regularized Lq-norm data mismatch term. The Lq-norm mismatch arises owing to the fact that the noise is sparse. The Lp-norm exploits the prior information that the image is sparse in a transform domain. The proposed methods have been used to reduce salt and pepper noise as well as random valued impulse noise. Peak signal to noise ratio and structural similarity index have been used to quantitatively evaluate the recovery results. A comparative study with existing IRN algorithm suggests the superiority of purposed algorithm. Our method also yields better results than the popular median filtering techniques used for denoising impulse noise.