Compressive sensing and adaptive direct sampling in hyperspectral imaging

In recent studies, superpixel segmentation has been integrated into hyperspectral (HS) image classification methods. However, the existing superpixel-based classification methods usually suffer from two serious problems. First, the accuracy and efficiency of current superpixel segmentation approaches cannot meet the demands of practical applications for HS images; second, conventional superpixel-based classification methods generally consider each generated superpixel as a unit for the image classification, which may help to reduce the computing time but result in a significant decrease of the classification accuracy. To solve the problems, we propose a fast region growing based superpixel segmentation (FRGSS) algorithm and a novel texture-adaptive superpixel integration strategy (TASIS) for the HS image classification. Experimental results on real Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) HS images demonstrate that the proposed FRGSS outperforms the state-of-the-art superpixel algorithm. In addition, the superiority of the TASIS is verified compared to the pixel-wise and the conventional superpixel-based classification methods.

Abstract. With the number of channels in the hundreds instead of in the tens Hyper spectral imagery possesses much richer spectral information than multispectral imagery. The increased dimensionality of such Hyper spectral data provides a challenge to the current technique for analyzing data. Conventional classification methods may not be useful without dimension reduction pre-processing. So dimension reduction has become a significant part of Hyper spectral image processing. This paper presents a comparative analysis of the efficacy of Haar and Daubechies wavelets for dimensionality reduction in achieving image classification. Spectral data reduction using Wavelet Decomposition could be useful because it preserves the distinction among spectral signatures. Daubechies wavelets optimally capture the polynomial trends while Haar wavelet is discontinuous and resembles a step function. The performance of these wavelets are compared in terms of classification accuracy and time complexity. This paper shows that wavelet reduction has more separate classes and yields better or comparable classification accuracy. In the context of the dimensionality reduction algorithm, it is found that the performance of classification of Daubechies wavelets is better as compared to Haar wavelet while Daubechies takes more time compare to Haar wavelet. The experimental results demonstrate the classification system consistently provides over 84% classification accuracy.

Hyperspectral image classification via nonlocal joint kernel sparse representation based on local covariance

Hyperspectral data compression has received considerable interest in recent years. Contrarily to the conventional compression schemes, which first acquire the full data set and then implement some compressing technique, compressive sensing (CS) acquires directly the compressed signal which will be later recovered on the ground station. The CS paradigm fits perfectly the requirements of onborad hyperspectral imaging systems in terms of energy, computing power, and bandwidth. By using CS in these systems, the amount of data acquired and transmitted to the ground stations is reduced and the bulk of the computation to infer the original data is carried out in the ground stations. In this paper, we present a new technique to perform CS of hyperspectral images (HSIs), which exploit the fact that HSIs admits a low dimensional linear representation. The proposed method is blind in the sense that linear representation is learned with low computational cost from the compressed measurements. Furthermore the proposed method is very light from the computational point of view and it can recover perfectly the original image in noise-free scenarios. The effectiveness of the proposed method is illustrated in both synthetic and real scenarios.

The purpose of this study is to develop hyperspectral imaging (HSI) for automatic detection of head and neck cancer cells on histologic slides. A compact hyperspectral microscopic system is developed in this study. Histologic slides from 15 patients with squamous cell carcinoma (SCC) of the larynx and hypopharynx are imaged with the system. The proposed nuclei segmentation method based on principle component analysis (PCA) can extract most nuclei in the hyperspectral image without extracting other sub-cellular components. Both spectra-based support vector machine (SVM) and patch-based convolutional neural network (CNN) are used for nuclei classification. CNNs were trained with both hyperspectral images and pseudo RGB images of extracted nuclei, in order to evaluate the usefulness of extra information provided by hyperspectral imaging. The average accuracy of spectra-based SVM classification is 68%. The average AUC and average accuracy of the HSI patch-based CNN classification is 0.94 and 82.4%, respectively. The hyperspectral microscopic imaging and classification methods provide an automatic tool to aid pathologists in detecting SCC on histologic slides.

Exploitation of the spectral capabilities of modern hyperspectral image demands efficient preprocessing and analyses methods. Analysts' choice of classifier and dimensionality reduction (DR) method and the harmony between them determine the accuracy of image classification. Multiple classifier system (MCS) has the potential to combine the relative advantages of several classifiers into a single image classification exercise for the hyperspectral image classification. In this paper, we propose an algorithmic extension of the MCS, named as dynamic classifier system (DCS), which exploits the context-based image and information class characteristics represented by multiple DR methods for hyperspectral image classification for land cover mapping. The proposed DCS algorithm pairs up optimal combinations of classifiers and DR methods specific to the hyperspectral image and performs image classifications based only on the identified combinations. Further, the impact of various trainable and nontrainable combination functions on the performance of the proposed DCS has been assessed. Image classifications were carried out on five multi-site airborne hyperspectral images using the proposed DCS and were compared with the MCS and SVM based supervised image classifications with and without DR. The results indicate the potential of the proposed DCS algorithm to increase the classification accuracy considerably over that of MCS or SVM supervised image classifications.

The compressed sensing (CS) theorem is a novel sampling approach that breaks through the conventional Nyquist sampling limit and brings a revolution in the field of signal processing. This article investigates the compression technique for CS hyperspectral images so as to illustrate the superiority provided by this new theorem. First, several comparative experiments are used to reveal that the drawback of prior compression techniques, designed for the data acquired by the conventional hyperspectral imaging system, is either low compression ratio or a waste of sampling resource. After a condensed analysis, we state that the CS theorem provides the probability of avoiding such defects. Then a straightforward scheme, which takes advantage of spectral correlation, is proposed to compress the CS hyperspectral images to reduce the data size further. Moreover, a flexible recovery strategy is designed to speed up the reconstruction of original bands from the corresponding CS images. The experimental results based on the actual hyperspectral images have demonstrated the efficiency of this proposed technique.

Spectral image clustering is an unsupervised classification method which identifies distributions of pixels using spectral information without requiring a previous training stage. The sparse subspace clustering-based methods (SSC) assume that hyperspectral images lie in the union of multiple low-dimensional subspaces. Using this, SSC groups spectral signatures in different subspaces, expressing each spectral signature as a sparse linear combination of all pixels, ensuring that the non-zero elements belong to the same class. Although these methods have shown good accuracy for unsupervised classification of hyperspectral images, the computational complexity becomes intractable as the number of pixels increases, i.e. when the spatial dimension of the image is large. For this reason, this paper proposes to reduce the number of pixels to be classified in the hyperspectral image, and later, the clustering results for the missing pixels are obtained by exploiting the spatial information. Specifically, this work proposes two methodologies to remove the pixels, the first one is based on spatial blue noise distribution which reduces the probability to remove cluster of neighboring pixels, and the second is a sub-sampling procedure that eliminates every two contiguous pixels, preserving the spatial structure of the scene. The performance of the proposed spectral image clustering framework is evaluated in three datasets showing that a similar accuracy is obtained when up to 50% of the pixels are removed, in addition, it is up to 7.9 times faster compared to the classification of the data sets without incomplete pixels.

Compressive sampling (CS), or Compressed Sensing , has generated a tremendous amount of excitement in the signal processing community. Compressive sampling, which involves non-traditional samples in the form of randomized projections, can capture most of the salient information in a signal with a relatively small number of samples, often far fewer samples than required using traditional sampling schemes. Adaptive sampling (AS), also called Active Learning , uses information gleaned from previous observations (e.g. , feedback) to focus the sampling process. Theoretical and experimental results have shown that adaptive sampling can dramatically outperform conventional (non-adaptive) sampling schemes. This paper compares the theoretical performance of compressive and adaptive sampling for regression in noisy conditions, and it is shown that for certain classes of piecewise constant signals and high SNR regimes both CS and AS are near optimal. This result is remarkable since it is the first evidence that shows that compressive sampling, which is non-adaptive, cannot be significantly outperformed by any other method (including adaptive sampling procedures), even in the presence of noise. The performance of CS schemes for signal detection is also investigated.© (2006) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

In this paper, a novel method based on wavelet transformation of functional data for accurate classification of hyperspectral images is proposed. The motivation of the proposed method is that in the hyperspectral images, the spectral curve of each pixel can be viewed as a function, mathematically. And it can make the best of the abundant spectral information of the original hyperspectral images. Based on this, we have proposed an effective hyperspectral image classification method which mainly used functional principal component analysis (FPCA) before. However, FPCA method focuses on the global features analysis of hyperspectral image data and may not well extract local structural features. In this paper, in order to better reflect the local structural features of hyperspectral images, we make use of the time-frequency localization characteristics of wavelet transformation to decompose hyperspectral images. Then, the sparse wavelet decomposition coefficients are classified by using extreme learning machine (ELM). Experimental results on two popular hyperspectral images show that our method outperforms some state-of-the-art hyperspectral image classification methods.

This paper presents a simulation method of hyper-spectral dynamic scene and image sequence for hyper-spectral equipment evaluation and target detection algorithm. Because of high spectral resolution, strong band continuity, anti-interference and other advantages, in recent years, hyper-spectral imaging technology has been rapidly developed and is widely used in many areas such as optoelectronic target detection, military defense and remote sensing systems. Digital imaging simulation, as a crucial part of hardware in loop simulation, can be applied to testing and evaluation hyper-spectral imaging equipment with lower development cost and shorter development period. Meanwhile, visual simulation can produce a lot of original image data under various conditions for hyper-spectral image feature extraction and classification algorithm. Based on radiation physic model and material characteristic parameters this paper proposes a generation method of digital scene. By building multiple sensor models under different bands and different bandwidths, hyper-spectral scenes in visible, MWIR, LWIR band, with spectral resolution 0.01μm, 0.05μm and 0.1μm have been simulated in this paper. The final dynamic scenes have high real-time and realistic, with frequency up to 100 HZ. By means of saving all the scene gray data in the same viewpoint image sequence is obtained. The analysis results show whether in the infrared band or the visible band, the grayscale variations of simulated hyper-spectral images are consistent with the theoretical analysis results.

Spectral features cannot effectively reflect the differences among the ground objects and distinguish their boundaries in hyperspectral image (HSI) classification. Multi-scale feature extraction can solve this problem and improve the accuracy of HSI classification. The Gaussian pyramid can effectively decompose HSI into multi-scale structures, and efficiently extract features of different scales by stepwise filtering and downsampling. Therefore, this paper proposed a Gaussian pyramid based multi-scale feature extraction (MSFE) classification method for HSI. First, the HSI is decomposed into several Gaussian pyramids to extract multi-scale features. Second, we construct probability maps in each layer of the Gaussian pyramid and employ edge-preserving filtering (EPF) algorithms to further optimize the details. Finally, the final classification map is acquired by a majority voting method. Compared with other spectral-spatial classification methods, the proposed method can not only extract the characteristics of different scales, but also can better preserve detailed structures and the edge regions of the image. Experiments performed on three real hyperspectral datasets show that the proposed method can achieve competitive classification accuracy.

The difficulties of obtaining sufficient labeled samples have always been one of the factors hindering deep learning models from obtaining high accuracy in hyperspectral image (HSI) classification. To reduce the dependence of deep learning models on training samples, meta learning methods have been introduced, effectively improving the classification accuracy in small sample set scenarios. However, the existing methods based on meta learning still need to construct a labeled source data set with several pre-collected HSIs, and must utilize a large number of labeled samples for meta-training, which is actually time-consuming and labor-intensive. To solve this problem, this paper proposes a novel unsupervised meta learning method with multiview constraints for HSI small sample set classification. Specifically, the proposed method first builds an unlabeled source data set using unlabeled HSIs. Then, multiple spatial-spectral multiview features of each unlabeled sample are generated to construct tasks for unsupervised meta learning. Finally, the designed residual relation network is used for meta-training and small sample set classification based on the voting strategy. Compared with existing supervised meta learning methods for HSI classification, our method can only utilize HSIs without any label for unsupervised meta learning, which significantly reduces the number of requisite labeled samples in the whole classification process. To verify the effectiveness of the proposed method, extensive experiments are carried out on 8 public HSIs in the cross-domain and in-domain classification scenarios. The statistical results demonstrate that, compared with existing supervised meta learning methods and other advanced classification models, the proposed method can achieve competitive or better classification performance in small sample set scenarios.

Abstract. Over the past decade hyper spectral (HS) image analysis has turned into one of the most powerful and growing technologies in the field of remote sensing. While HS images cover large area at fine spectral resolution, their spatial resolutions are often too coarse for the use in various applications. Hence improving their resolution has a high payoff. This paper presents a novel approach for super-resolution (SR) of HS images using compressive sensing (CS). Besides ill-posedness of SR problem, the main challenge in HS super-resolution is to preserve spectral contents among all bands while increasing their spatial resolutions. In this work, we first obtain an initial estimate of the super-resolution on a reduced dimension HS data. The HS observations of different wavelengths are represented as linear combination of smaller number of basis image planes (BIPs) using principal component analysis (PCA). The novelty of our approach lies in using CS based approach to super-resolve the most informative PCA transformed image representing highest spectral variance (i.e. the first principal component). Our approach uses low and high spatial resolution dictionaries of patches generated by random sampling of raw patches of PCA transformed images that are generated using the training images having similar statistical properties. Using the sparsity constraint, low resolution test patch is represented as a sparse linear combination of relevant dictionary elements adaptively, that leads to initial estimate of super-resolved PCA image having maximum spectral variability. Since SR is an ill-posed problem, we obtain the final solution using a regularization framework considering the sparse coefficients obtained by the CS approach and the autoregressive (AR) parameters obtained from the initial estimate. The remaining PCA images are up-scaled using regularization, considering the same AR parameters which were obtained from super-resolved PCA image having maximum spectral variability. Experiments are conducted on real HS images collected by Airborne Visible/Infrared Imaging Spectrometer (AVIRIS). Visual inspections and quantitative comparison confirm that our method enhances spatial information without introducing significant spectral distortion.

Abstract. Hyperspectral images (HSIs) contain hundreds of spectral bands, providing high-resolution spectral information pertaining to the Earth’s surface. Additionally, abundant spatial contextual information can also be obtained simultaneously from a HSI. To characterize the properties of ground objects, classification is the most widely-used technology in the field of remote sensing, where each pixel in a HSI is assigned to a pre-defined class. Over the past decade, deep learning has attracted increasing attention in the machine-learning and computer-vision domains, due to its favourable performances for various types of tasks, and it has been successfully introduced to the remote-sensing community. Instead of utilizing the shallow features within in a given image, which is the approach that is generally adopted in other conventional classification methods, deep-learning algorithms can extract hierarchical features from raw HSI data. Within the deep-learning framework, recurrent neural networks (RNNs), which are able to encode sequential features, have exhibited promising capabilities and have achieved encouraging performances, especially for the natural-language processing and speech-recognition communities. As multi-temporal remote-sensing images can be readily obtained from increasing numbers of satellite and unmanned aircraft systems, and since analysis of such multi-temporal data comprises a critical issue within numerous research subfields, including land-cover and land-change analyses, and land-resource management, RNNs have been applied in recent studies in order to extract temporal sequential features from multi-temporal remote-sensing images for the purpose of image classification. Apart from using multi-temporal image datasets, RNNs can also be utilized on a single image, where the spectral feature/band of each individual pixel can be taken as a sequential feature for the input layer of RNNs. However, the application of such sequential feature extraction that relies on a single image still needs to be further investigated since applying RNNs to spectral bands will directly introduce more parameters that need to be optimized, consequently increasing the total training time.In this study, we propose a novel RNN-based HSI classification framework. In this framework, unlabelled pixels obtained from a single image are considered when constructing sequential features. Two spatial similarity measurements, referred to as pixel-matching and block-matching, respectively, are employed to extract pixels that are “similar” to the target pixel. Then, the sequential feature of the target pixel is constructed by exploiting several of the most “similar” pixels and ordering them based on their similarities to the target pixel. The aforementioned two schemes are advantageous, as unlabelled pixels within the given HSI are taken into consideration for similarity measurement and sequential feature construction for the RNN model. Moreover, the block-matching scheme also takes advantage of spatial contextual information, which has been widely utilized in spatial-spectral-based HSI classification methods. To evaluate the proposed methods, two benchmark HSIs are used, including a HSI collected over Pavia University, Italy by the airborne Reflective Optics System Imaging Spectrometer (ROSIS) sensor, and an image acquired over the Salinas Valley, California, USA via the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) sensor. Spatio-temporally coincident ground-reference data accompanies each of these respective HSIs. In addition, the proposed methods are compared with three state-of-the-art algorithms, including support vector machine (SVM), the 1-dimensional convolutional neural network (1DCNN), and the 1-dimensional RNN (1DRNN).Experimental results indicate that our proposed methods achieve markedly better classification performance compared with the baseline algorithms on both datasets. For example, for the Pavia University image, the block-matching based RNN achieves the highest overall classification accuracy, with 94.32% accuracy, which is 9.87% higher than the next most accurate algorithm of the aforementioned three baseline methods, which in this case is the 1DCNN, with 84.45% overall accuracy. More specifically, the block-matching method performs better than the pixel-matching method in terms of both quantitative and qualitative assessments. Based on visual assessment/interpretation of the classification maps, it is apparent that “salt-and-pepper” noise is markedly alleviated; with block-matching, smoother classified images are generated compared with pixel-matching-based methods and the three baseline algorithms. Such results demonstrate the effectiveness of utilizing spatial contextual information in the similarity measurement.