Xiurui Geng

A Fast Volume-Gradient-Based Band Selection Method for Hyperspectral Image

In this paper, a subtle relationship is found between the volume of a subsimplex and the volume gradient of a simplex with respect to hyperspectral images. By using this relationship, we propose an efficient band selection method, namely, the volume-gradient-based band selection (VGBS) method. The VGBS method is an unsupervised method, which tries to remove the most redundant band successively. Interestingly, the VGBS method can find the most redundant band based only on the gradient of volume instead of calculating the volumes of all subsimplexes. Experiments on simulated and real hyperspectral data verify the efficiency of the proposed method.

A New Band Selection Method for Hyperspectral Image Based on Data Quality

Most unsupervised band selection methods take the information of bands into account, but few of them pay attention to the quality of bands. In this paper, by combining idea of noiseadjusted principal components (NAPCs) with a state-of-art band selection method [maximum determinant of covariance matrix (MDCM)], we define a new index to quantitatively measure the quality of the hyperspectral data cube. Both signal-to-noise ratios (SNRs) and correlation of bands are simultaneously considered in . Based on the new index defined in this article, we propose an unsupervised band selection method called minimum noise band selection (MNBS). Taking the quality (Q) of the data cube as selection criterion, MNBS tries to find the bands with both high SNRs and low correlation (high ). The subset selection method, sequential backward selection (SBS), is used in MNBS to improve the search efficiency. Some comparative experiments based on simulated as well as real hyperspectral data are conducted to evaluate the performance of MNBS in this study. The experimental results show that the bands selected by MNBS are always more effective than those selected by other methods in terms of classification.

Exemplar Component Analysis: A Fast Band Selection Method for Hyperspectral Imagery

How to find the representative bands is a key issue in band selection for hyperspectral data. Very often, unsupervised band selection is associated with data clustering, and the cluster centers (or exemplars) are considered ideal representatives. However, partitioning the bands into clusters may be very time-consuming and affected by the distribution of the data points. In this letter, we propose a new band selection method, i.e., exemplar component analysis (ECA), aiming at selecting the exemplars of bands. Interestingly, ECA does not involve actual clustering. Instead, it prioritizes the bands according to their exemplar score, which is an easy-to-compute indicator defined in this letter measuring the possibility of bands to be exemplars. As a result, ECA is of high efficiency and immune to distribution structures of the data. The experiments on real hyperspectral data set demonstrate that ECA is an effective and efficient band selection method.

A New Sparsity-Based Band Selection Method for Target Detection of Hyperspectral Image

Band selection (BS) plays an important role in the dimensionality reduction of hyperspectral data. However, as to the existing BS methods, few are specially designed for target detection. In this letter, we combine the target detection and BS process together and put forward a new BS method for target detection, named least absolute shrinkage and selection operator (LASSO)-based BS (LBS). Interestingly, by using a linear regression model with L1 regularization (LASSO model), LBS transforms the discrete BS problem into the continuous optimization problem, which cannot only avoid the complicated subset selection process but also evaluate the importance of all the bands simultaneously. The experiments on real hyperspectral data demonstrate that LBS is a very effective BS method for target detection.

CEM: More Bands, Better Performance

Target detection has recently drawn considerable interest in hyperspectral image processing. People tend to exclude corrupted or badly damaged bands before applying the target detection algorithm to the data for better detection results. In this letter, it is proved that adding any band independent of the original image, even a noisy band, would be always beneficial to the performance of constrained energy minimization in terms of output energy. Finally, several tests are conducted to further justify our viewpoint.

A Fast Endmember Extraction Algorithm Based on Gram Determinant

In the field of endmember extraction, most methods involve calculating the volume of simplex in high-dimensional space. Two different simplex volume formulas are used in these methods. One requires dimensionality reduction (DR); therefore, it may result in loss of the information of targets classes with a low priori probability, such as that used in N-FINDR. The other one, which is based on Gram determinant, avoids DR but is time consuming. In this letter, we explain a recursion rule of the calculation for the second simplex volume. Based on that rule, this letter presents a fast endmember extraction algorithm named as Fast Gram Determinant based Algorithm (FGDA). The theoretical analysis and experiments on both simulated and real hyperspectral data demonstrate that, compared to other volume-based methods, FGDA can greatly reduce the computational complexity of endmember extraction.

A new volume formula for a simplex and its application to endmember extraction for hyperspectral image analysis

Based on the geometric properties of a simplex, endmembers can be extracted automatically from a hyperspectral image. To avoid the shortcomings of the N-FINDR algorithm, which requires the dimensions of the data to be one less than the number of endmembers needed, a new volume formula for the simplex without the requirement of dimension reduction is presented here. We demonstrate that the N-FINDR algorithm is a special case of the new method. Moreover, whether the null vector is included as an endmember has an important effect on the final result of the endmember extraction. Finally, we compare the new method with previous methods for endmember extraction of hyperspectral data collected by the Advanced Visible and Infrared Imaging Spectrometer (AVIRIS) over Cuprite, Nevada.

A New Endmember Generation Algorithm Based on a Geometric Optimization Model for Hyperspectral Images

This letter presents a new endmember generation method, which is called the geometric optimization model (GOM). The algorithm exploits the following fact: an L-dimensional (L-D) simplex can be divided into L + 1 L-D smaller simplexes by any point within the simplex, and the sum of the volumes of the L + 1 smaller simplexes is equal to the volume of the simplex. Based on this geometrical property, we propose a new objective function for endmember generation, whose variable only includes the mixing matrix. As a result, all the problems caused by the abundance matrix can be avoided. Experiments using both simulated and real hyperspectral data show that the GOM is effective in searching the optimal solution.

A High-Order Statistical Tensor Based Algorithm for Anomaly Detection in Hyperspectral Imagery

Recently, high-order statistics have received more and more interest in the field of hyperspectral anomaly detection. However, most of the existing high-order statistics based anomaly detection methods require stepwise iterations since they are the direct applications of blind source separation. Moreover, these methods usually produce multiple detection maps rather than a single anomaly distribution image. In this study, we exploit the concept of coskewness tensor and propose a new anomaly detection method, which is called COSD (coskewness detector). COSD does not need iteration and can produce single detection map. The experiments based on both simulated and real hyperspectral data sets verify the effectiveness of our algorithm.

Improving the Accuracy of the Water Surface Cover Type in the 30 m FROM-GLC Product

The finer resolution observation and monitoring of the global land cover (FROM-GLC) product makes it the first 30 m resolution global land cover product from which one can extract a global water mask. However, two major types of misclassification exist with this product due to spectral similarity and spectral mixing. Mountain and cloud shadows are often incorrectly classified as water since they both have very low reflectance, while more water pixels at the boundaries of water bodies tend to be misclassified as land. In this paper, we aim to improve the accuracy of the 30 m FROM-GLC water mask by addressing those two types of errors. For the first, we adopt an object-based method by computing the topographical feature, spectral feature, and geometrical relation with cloud for every water object in the FROM-GLC water mask, and set specific rules to determine whether a water object is misclassified. For the second, we perform a local spectral unmixing using a two-endmember linear mixing model for each pixel falling in the water-land boundary zone that is 8-neighborhood connected to water-land boundary pixels. Those pixels with big enough water fractions are determined as water. The procedure is automatic. Experimental results show that the total area of inland water has been decreased by 15.83% in the new global water mask compared with the FROM-GLC water mask. Specifically, more than 30% of the FROM-GLC water objects have been relabeled as shadows, and nearly 8% of land pixels in the water-land boundary zone have been relabeled as water, whereas, on the contrary, fewer than 2% of water pixels in the same zone have been relabeled as land. As a result, both the user’s accuracy and Kappa coefficient of the new water mask (UA = 88.39%, Kappa = 0.87) have been substantially increased compared with those of the FROM-GLC product (UA = 81.97%, Kappa = 0.81).