Zhenyu An

Hyperspectral image fusion by the similarity measure-based variational method

Hyperspectral remote sensing is widely used in many fields suchas agriculture, military detection, mineral exploration, and so on. Hyperspectral data has very high spectral resolution, but much lower spatial resolution than the data obtained by other types of sensors. The low spatial resolution restrains its wide applications. On the contrary, we easily obtain images with high spatial resolution but insufficient spectral resolution (like panchromatic images). Naturally, people expect to obtain images that have high spatial and spectral resolution at the same time by the hyperspectral image fusion. In this paper, a similarity measure-based variational method is proposed to achieve the fusion process. The main idea is to transform the image fusion problem to an optimization problem based on the variational model. We first establish a fusion model that constrains the spatial and spectral information of the original data at the same time, then use the split bregman iteration to obtain the final fused data. Also, we analyze the convergence of the method. The experiments on the synthetic and real data show that the fusion method preserves the information of the original images efficiently, especially on the spectral information.

Scene Learning for Cloud Detection on Remote-Sensing Images

Cloud detection plays a major role for remote-sensing image processing. To accomplish the task, a novel automatic supervised approach based on the “scene-learning” scheme is proposed in this paper. Scene learning aims at training and applying a cloud detector on the whole image scenes. The cloud detector herein is a special classifier that is used to separate clouds from the backgrounds. Concretely, scene learning regards each pixel of scenes in training image as a sample, and uses it to train a cloud detector. Accordingly, the detecting process is also implemented on each pixel of testing image using the trained detector. Generally, scene-learning scheme contains two modules: feature data simulating and cloud detector learning and applying. We first simulate a kind of cubic structural data (also named feature data) by stacking different fundamental image features, including color, statistical information, texture, and structure. Such data synthesize different image features, and it is used for cloud detector training and applying. Cloud detector is designed based on minimizing the residual error between the feature data and its labels. The detector is easy to be trained because of its closed-form. Applying the detector and some necessary cloud refinement methods to the testing images, we could finally detect clouds. We also theoretically analyze the influence of feature number and prove that more features lead to better performance of scene learning under certain circumstance. Comparisons of qualitative and quantitative analyses of the experimental results are implemented. Results indicate the efficacy of the proposed method.

Nonnegative matrix factorization for hyperspectral unmixing using prior knowledge of spectral signatures

Hyperspectral unmixing is a process aiming at identifying the constituent materials and estimating the corresponding fractions from hyperspectral imagery of a scene. Nonnegative matrix factorization (NMF), an effective linear spectral mixture model, has been applied in hyperspectral unmixing during recent years. As the data of hyperspectral imagery analyzed deeper, prior knowledge of some signatures in the scene could be available. In several scenes such as mining areas, a few surface substances like copper and iron are easy to identify through field investigation. Thus, their spectral signatures can be used as prior knowledge to unmix hyperspectral data. In such a context, we propose an NMF based framework for hyperspectral unmixing using such prior knowledge, referred to as NMFupk. Specifically, our algorithm supposes that some spectral signatures in the scene are known and then utilizes the prior knowledge of the spectral signatures to unmix the hyperspectral data. In a series of experiments, we test NMFupk and NMF without prior knowledge on both synthetic and real data. Results achieved demonstrate the efficacy of the proposed algorithm.

A Novel Spectral-Unmixing-Based Green Algae Area Estimation Method for GOCI Data

Geostationary Ocean Color Imager (GOCI) data have been widely used in the detection and area estimation of green algae blooms. However, due to the low spatial resolution of GOCI data, pixels in an image are usually “mixed,” which means that the region a pixel covers may include many different materials. Traditional index-based methods can detect whether there are green algal blooms in each pixel, whereas it is still challenging to determine the proportion that green algae blooms occupy in a pixel. In this paper, we propose a novel subpixel-level area estimation method for green algae blooms based on spectral unmixing, which can not only detect the existence of green algae but also determine their proportion in each pixel. A fast endmember extraction method is proposed to automatically calculate the endmember spectral matrix, and the abundance map of green algae which could be regarded as the area estimation is obtained by nonnegatively constrained least squares. This new fast endmembers extraction technique outperforms the classical N-FINDR method by applying two models: candidates location and distance-based vertices determination. In the first model, we propose a medium-distance-based candidates location strategy, which could reduce the searching space during vertices selection. In the second model, we replace the simplex volume measure with a more simple distance measure, thus complex matrix determinant calculation is avoided. We have theoretically proven the equivalency of volume and distance measure. Experiments on GOCI data and synthetic data demonstrate the superiority of the proposed method compared with some state-of-art approaches.

Panchromatic image processing using hyperspectral unmixing method

In the paper, we consider the probability of applying hyper-spectral image (HSI) processing methods to panchromatic images (PIs), which is a novel yet crucial issue for further analyses. To achieve the purpose, we propose an effective approach for handling PI with HSI unmixing methods. In the approach, HSI simulating process is first implemented to obtain a synthetic HSI from PI. After that, a hyperspectral unmixing algorithm, vertex component analysis, is then applied to extract endmembers that comprise the vertices of the data simplex. Meanwhile, we calculate abundances of HSI by employing least squares method. We will see that the unmixing results, namely endmembers and abundances, can be used for target detection and other applications. Different objects such as ships and cars were successfully extracted from the backgrounds, which demonstrates the efficacy of the proposed approach.

A novel unsupervised approach to discovering regions of interest in traffic images

Analyzing image of traffic scenes plays a major role in intelligent transportation systems. Regions of interest, including traffic signs, vehicles or some other man-made objects, largely attract drivers? attention. With different prior knowledge, conventional approaches generally define and build dedicated detectors to each class of such regions. In contrast, this paper focuses on explaining what regions in traffic images can be of interest, which is a critical problem yet neglected before. Instead of pre-defining the detectors, a computational model based on an unsupervised way is proposed. The core idea is to simulate an image with multiple bands from the given traffic image by stacking the spatial information. Our study shows that the distribution of such data can be captured by a simplex in a linear subspace, and each data point can be represented by a linear reconstruction over the set of vertices of the simplex. An effective method to identify the simplex vertices is proposed. These simplex vertices actually comprise the core elements in the regions of interest, as physically they correspond to regions with saturated colors. Comparisons of the proposed approach and conventional methods on computational complexity and practical extensive experiments are implemented. The results validate and show the efficacy of the proposed approach. HighlightsThis paper focuses on explaining what regions in traffic images can be of interest.A computational model based on an unsupervised way is proposed.Our study shows that the distribution of such data can be captured by a simplex in a linear subspace.An effective method to identify the simplex vertices is proposed.The results validate and show the efficacy of the proposed approach.

Hyperspectral unmixing using non-negative matrix factorization with automatically estimating regularization parameters

Hyperspectral unmixing is a process by which pixel spectra in a scene are decomposed into constituent materials and their corresponding fractions. Nonnegative matrix factorization (NMF) is a method recently developed to deal with matrix factorization. This paper proposes a hyperspectral unmixing algorithm using auto-NMF based on the L-curve theory. It is an approach to automatically estimate regularization parameters, which are manually chosen subjectively and difficultly in the traditional regularized non-negative matrix factorization (RNMF). We experiment traditional algorithms and auto-NMF on the synthetic data, better results are obtained from auto-NMF, indicating it is an effective technique for hyperspectral unmixing.