Supervised Band Selection Using Local Spatial Information for Hyperspectral Image

In this paper, a contextual Support Vector Machine (SVM) technique based on the principle of Hilbert Space Embedding (HSE) of a local hyperspectral data distribution into an Reproducing Kernel Hilbert Space (RKHS) is proposed to optimally exploit the spectral and local spatial information of the hyperspectral image. The idea of embedding is to map hyperspectral pixels in a local neighborhood into a single point in the RKHS that can uniquely represent those pixels collectively. Previously, the authors have employed an HSE called empirical mean map to build the contextual SVM. In this work, a weighted empirical mean map is utilized to exploit the similarities and variation in the local spatial information. For every pixel, a small set of the neighboring pixels in a hyperspectral image are mapped into an RKHS induced by a certain kernel (Eg. Gaussian RBF kernel) and then, the embedded point of these group of pixels is obtained by calculating the weighted empirical mean of these mapped points. The weights are determined based on the distance between the pixel in consideration and its neighbors. An SVM separating hyperplane is built to maximize the margin between classes formed by weighted empirical means. The proposed technique showed significant improvement over the existing contextual and composite kernels on two hyperspectral image data sets.

MIMN-DPP: Maximum-information and minimum-noise determinantal point processes for unsupervised hyperspectral band selection

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] The possibilistic c-means (PCM) was developed as an extension of the fuzzy c-means (FCM) clustering algorithm by abandoning the membership sum-to-one constraint. In the PCM, each cluster is independent of the other clusters and can be processed separately. Because of this separability, the sequential possibilistic one-mean (SP1M) was proposed to find clusters sequentially by running the possibilistic one-mean (P1M) c times. One critical problem in both PCM and SP1M is how to determine the parameter [subscript]. The sequential possibilistic one-mean with adaptive eta (SP1M-AE) was developed to allow [subscript] to change during iterations. In this thesis, a new dynamic adaptation mechanism for the parameter [subscript] in each cluster is inserted into SP1M. The resultant algorithm, called the sequential possibilistic one-mean with dynamic eta (SP1M-DE) is shown to provide superior performance over PCM, SP1M, and SP1M-AE in determining correct clustering results. In this thesis, the family of the SP1M clustering algorithm is also extended to combine local spatial information in an image. The resultant algorithm, called sequential possibilistic local information one-mean (SPLI1M) is shown to have a better performance on image segmentation over the fuzzy local information c-means (FLICM), the possibilistic local information c-means (PLICM), and the family of SP1M without combing local spatial information in image.

Band selection algorithm is most important in data dimension reduction of hyperspectral image. There are many algorithms of band selection, but there are only few methods to do algorithm evaluation. A method is put forward in this paper to evaluate the band selection algorithm of hyperspectral image. The amount of information, brightness, image contrast and definition are defined as 4 indexes to measure deferent data fusion based on various band selection results. Based on the measurement, the evaluation of band selection algorithm is realized. In the paper, the evaluation method is used in the compare of 4 common band selection algorithms, the result of measurement is analyzed and the feasibility is verified.

Hyperspectral remote sensing sensors have the ability to capture a wide range of spectrum of ground objects with hundreds to thousands of bands. The obtained hyperspectral images contain more detailed spectral information than conventional panchromatic or color images. However, redundant and noisy data could also be introduced into the images and may impair further processing of the data. Removing the redundancy and noisy bands becomes one of the most important preprocessing steps of hyperspectral imaging. As a feature selection method, an attention mechanism has been successfully used in computer vision and natural language processing to enable algorithms to concentrate on the most significant data. In this article, we propose a band attention network for hyperspectral image classification. This network can automatically learn to attend to the desired band set that maximizes the classification accuracy. With careful design, the band attention framework can undertake both band weighting and band selection tasks. When working in the band weighting mode, the proposed band attention framework can learn a band weighting vector that models the relationship between all the bands. For band selection, our network can be adapted as two different types of band selection networks that select the significant bands and discard the useless bands. In both tasks, the attention and classification can be learned end to end. The experimental results prove the effectiveness and advantages of the proposed work.

As fuzzy c-means clustering (FCM) algorithm is
\nsensitive to noise, local spatial information is often introduced
\nto an objective function to improve the robustness of the FCM
\nalgorithm for image segmentation. However, the introduction of
\nlocal spatial information often leads to a high computational
\ncomplexity, arising out of an iterative calculation of the distance
\nbetween pixels within local spatial neighbors and clustering
\ncenters. To address this issue, an improved FCM algorithm
\nbased on morphological reconstruction and membership filtering
\n(FRFCM) that is significantly faster and more robust than FCM,
\nis proposed in this paper. Firstly, the local spatial information
\nof images is incorporated into FRFCM by introducing morphological
\nreconstruction operation to guarantee noise-immunity
\nand image detail-preservation. Secondly, the modification of
\nmembership partition, based on the distance between pixels
\nwithin local spatial neighbors and clustering centers, is replaced
\nby local membership filtering that depends only on the spatial
\nneighbors of membership partition. Compared to state-of-theart
\nalgorithms, the proposed FRFCM algorithm is simpler and
\nsignificantly faster, since it is unnecessary to compute the distance
\nbetween pixels within local spatial neighbors and clustering
\ncenters. In addition, it is efficient for noisy image segmentation
\nbecause membership filtering are able to improve membership
\npartition matrix efficiently. Experiments performed on synthetic
\nand real-world images demonstrate that the proposed algorithm
\nnot only achieves better results, but also requires less time than
\nstate-of-the-art algorithms for image segmentation.

Different crops, as well as the same crop at different growth stages, display distinct spectral and spatial characteristics in hyperspectral images (HSIs) due to variations in their chemical composition and structural features. However, the narrow bandwidth and closely spaced spectral channels of HSIs result in significant data redundancy, posing challenges to crop identification and classification. Therefore, the dimensionality reduction in HSIs is crucial. Band selection as a widely used method for reducing dimensionality has been extensively applied in research on crop identification and mapping. In this paper, a crop superpixel-based affinity propagation (CS-AP) band selection method is proposed for crop identification and mapping in agriculture using HSIs. The approach begins by gathering crop superpixels; then, a spectral band selection criterion is developed by analyzing the variations in the spectral and spatial characteristics of crop superpixels. Finally, crop identification bands are determined through an efficient clustering approach, AP. Two typical agricultural hyperspectral data sets, the Salinas Valley data set and the Indian Pines data set, are selected for validation, each containing 16 crop classes, respectively. The experimental results show that the proposed CS-AP method achieves a mapping accuracy of 92.4% for the Salinas Valley data set and 88.6% for the Indian Pines data set. When compared to using all bands, two unsupervised band selection techniques, and three semi-supervised band selection techniques, the proposed method outperforms others with an improvement of 3.1% and 4.3% for the Salinas Valley and Indian Pines data sets, respectively. Indicate that the proposed CS-AP method achieves superior mapping accuracy by selecting fewer bands with greater crop identification capability compared to the other band selection methods. This research’s significant results demonstrate the potential of this approach in precision agriculture, offering a more cost-effective and timely solution for large-scale crop mapping and monitoring in the future.

Band selection for hyperspectral images is an effective technique to mitigate the curse of dimensionality. A variety of band selection methods have been suggested in the past. This paper presents a novel band prioritization based on impurity function (IF) for the band selection of hyperspectral images. The proposed IF band selection (IFBS) is incorporated with particle swarm optimization (PSO) band selection which has been developed to effectively group highly correlated bands of hyperspectral images into high corrected modules. It uses a particle swarm optimization scheme, which is a well-known method to solve the optimization problems, to develop an effective feature extraction algorithm for hyperspectral imagery. After PSO method is applied to the band reduction of hyperspectral images, the proposed IFBS is applied to enhance the efficiency of band selection. The propose method is evaluated by MODIS/ASTER airborne simulator (MASTER) for land cover classification during the Pacrim II campaign. The performance of IFBS is validated by the supervised k-nearest neighbor (KNN) classifier. Experimental results demonstrate that the proposed IFBS approach is an effective method for dimensionality reduction and feature extraction. Compared to other band selection methods, IFBS can effectively select the most significant bands for the image classification of hyperspectral images.

This paper presents a promising band selection algorithm, known as particle swarm optimization/impurity function class overlapping (PSO/IFCO) method, which adopts a novel multiple attribute decision making (MADM) model approach to the hyperspectral remote sensing images. The proposed MADM-based PSO/IFCO method can be divided into two steps: 1) PSO algorithm and 2) the IFCO scheme. With PSO band selection algorithm, the highly correlated bands of hyperspectral imagery can first be grouped into band modules, known as greedy modular eigenspace (GME), to coarsely reduce high-dimensional datasets in the first step. The more highly correlated small modules are further constructed with the statistics of impurity weights calculated by IFCO scheme in the second step. These statistics results of impurity weights are used to finely select the most important feature bands from the hyperspectral imagery. The proposed MADM-based PSO/IFCO makes use of the correlation coefficients matrix to cluster the highly correlated bands together and obtain GME in the first step. More specifically, we use the analytic hierarchy process (AHP) model, which is the most suitable implementation of MCDM for proposed method, to examine hierarchically the relations among different GME modules with the impurity weighted by IFCO in the second step. Finally, by accommodating the statistics of impurity weights, the proposed MADM-based PSO/IFCO method can effectively select the most representative features for hyperspectral band selection and reduction. The effectiveness of the proposed method is evaluated by MASTER and AVIRIS hyperspectral images. The experimental results demonstrate that the proposed method can not only enhance the high dimension reduction rate, but also offer a satisfactory classification performance.

Multi-agent deep reinforcement learning for hyperspectral band selection with hybrid teacher guide

With the advancement in technology, hyperspectral images have potential applications in the field of remote sensing due to their high spectral resolution. Despite the hyperspectral image providing abundant information, its analysis suffers from the problem of high dimensionality. Hence, Dimensionality Reduction (DR) is an essential task in all hyperspectral image analysis. Band Selection, which is one of the DR techniques, is still a challenging issue even though many algorithms have been developed. To provide remedy for this issue, this paper explores a novel approach for band selection using K-means clustering on statistical feature in hyperspectral images. The proposed method of clustering based band selection for DR is simple and accurate. A reliable estimate of number of bands to be selected is provided by Virtual Dimensionality (VD). Informative bands preserving maximum information are selected based on the statistical feature, the variance using K-means Clustering technique. Further, our proposed work involves the utilization of the effectiveness of Simplex Growing Algorithm (SGA) on endmember extraction in association with clustering based band selection. Using Fully Constrained Least Squares (FCLS) method, abundance fraction is estimated based on endmember signatures, which are derived using Endmember Extraction Algorithm (EEA). The proposed work is investigated and compared with that of N-FINDR and Vertex Component Analysis (VCA) algorithms. The performance of the proposed algorithm is evaluated using Root Mean Square Error (RMSE), Spectral Angle Distance (SAD) and computation time. Experimental results show that the proposed clustering based band selection with SGA endmember extraction algorithm reduces the average SAD by 8 to 10 % and the average RMSE by nearly 1 %, compared to that of N-FINDR and VCA algorithms. In terms of computation time, the proposed band selection based DR with SGA algorithm is seven times faster than conventional transform based DR with SGA algorithm.

Band selection (BS) is an effective technique of dimensionality reduction in hyperspectral images (HSIs), which can solve the problems of high computational complexity and information redundancy and is helpful for the classification of HSIs. Since many hyperspectral scenes have limited label samples, semisupervised BS methods have attracted much attention. In this paper, an improved graph-based semisupervised BS method is proposed. The within-class and between-class graphs are firstly constructed by utilizing the information of the labeled and unlabeled samples, then Fisher's criteria combined with sparsity regularization (FCS) is designed to obtain the optimal projection matrix, which is used to select the representative bands. The proposed algorithm performs BS while preserving the local manifold structure of data. Experimental results on three hyperspectral data sets show that the proposed BS algorithm has good performance in selecting representative bands for classification.

Hyperspectral band selection (BS) methods relying solely on volume-based criteria often fail to exclude noisy bands and may yield redundant subsets with limited representativeness. To address these limitations, we propose two novel BS algorithms that integrate canonical correlation analysis (CCA) and projection energy measures with determinant-based volume criterion. By modelling the correlation structure between selected and unselected bands, we effectively suppress the inclusion of noisy or outlier bands and enhance both the representativeness and independence of the chosen subsets. Furthermore, we introduce complexity reduction techniques, including the simplified computation of canonical correlation coefficients and recursive updates of inverse matrices, to facilitate efficient implementation for large-scale datasets. Experimental results on five real-world hyperspectral datasets demonstrate that the proposed methods achieve higher classification accuracies, greater robustness against noise, and significantly lower computational costs compared to traditional volume-based and linear algebra-based BS algorithms. These findings underscore the effectiveness and practical value of integrating correlation structure into volume-based BS frameworks.

In this paper, well-known traditional band selection methods which are used in hyperspectral imaging, namely, Maximum-Variance Principal Component Analysis (MVPCA), Maximum-SNR Principal Component Analysis (MSNRPCA), k-means, k-medoids, and recently proposed Automatic Band Selection (ABS) and Band Column Selection (BCS) approaches are compared. To assess the band selection performance of the methods, the change of classification performance by the number of selected bands is used. Performances of the methods are evaluated on three hyperspectral data sets and obtained results are compared in this paper.

Greedy modular eigenspaces (GME) has been developed for the band selection of hyperspectral images (HSI). GME attempts to greedily select uncorrelated feature sets from HSI. Unfortunately, GME is hard to find the optimal set by greedy operations except by exhaustive iterations. The long execution time has been the major drawback in practice. Accordingly, finding an optimal (or near-optimal) solution is very expensive. In this study we present a novel parallel mechanism, referred to as parallel particle swarm optimization (PPSO) band selection, to overcome this disadvantage. It makes use of a new particle swarm optimization scheme, a well-known method to solve the optimization problems, to develop an effective parallel feature extraction for HSI. The proposed PPSO improves the computational speed by using parallel computing techniques which include the compute unified device architecture (CUDA) of graphics processor unit (GPU), the message passing interface (MPI) and the open multi-processing (OpenMP) applications. These parallel implementations can fully utilize the significant parallelism of proposed PPSO to create a set of near-optimal GME modules on each parallel node. The experimental results demonstrated that PPSO can significantly improve the computational loads and provide a more reliable quality of solution compared to GME. The effectiveness of the proposed PPSO is evaluated by MODIS/ASTER airborne simulator (MASTER) HSI for band selection during the Pacrim II campaign.