Hyperspectral Band Selection via Optimal Combination Strategy

Li, Shuying; Peng, Baidong; Fang, Long; Li, Qiang

doi:10.3390/rs14122858

Cited by 11 publications

(3 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In CFNR, considering the sensitivity of FCM to noise [ 39 , 40 ], we aim to select the target band subset via a method that can effectively reduce the effect of noise on the band selection. Specifically, to obtain the target band subset based on the clustering results, the proposed CFNR approach adopts the information entropy-based method [ 41 , 42 ], using which the band that contains the maximum amount of information in each cluster is selected as the representative band. This method is based on the assumption that bands should be selected based on the amount of information contained in the band.…”

Section: Proposed Methodologymentioning

confidence: 99%

Joint Learning of Correlation-Constrained Fuzzy Clustering and Discriminative Non-Negative Representation for Hyperspectral Band Selection

Wang

2023

Sensors

View full text Add to dashboard Cite

Hyperspectral band selection plays an important role in overcoming the curse of dimensionality. Recently, clustering-based band selection methods have shown promise in the selection of informative and representative bands from hyperspectral images (HSIs). However, most existing clustering-based band selection methods involve the clustering of original HSIs, limiting their performance because of the high dimensionality of hyperspectral bands. To tackle this problem, a novel hyperspectral band selection method termed joint learning of correlation-constrained fuzzy clustering and discriminative non-negative representation for hyperspectral band selection (CFNR) is presented. In CFNR, graph regularized non-negative matrix factorization (GNMF) and constrained fuzzy C-means (FCM) are integrated into a unified model to perform clustering on the learned feature representation of bands rather than on the original high-dimensional data. Specifically, the proposed CFNR aims to learn the discriminative non-negative representation of each band for clustering by introducing GNMF into the model of the constrained FCM and making full use of the intrinsic manifold structure of HSIs. Moreover, based on the band correlation property of HSIs, a correlation constraint, which enforces the similarity of clustering results between neighboring bands, is imposed on the membership matrix of FCM in the CFNR model to obtain clustering results that meet the needs of band selection. The alternating direction multiplier method is adopted to solve the joint optimization model. Compared with existing methods, CFNR can obtain a more informative and representative band subset, thus can improve the reliability of hyperspectral image classifications. Experimental results on five real hyperspectral datasets demonstrate that CFNR can achieve superior performance compared with several state-of-the-art methods.

show abstract

Section: Proposed Methodologymentioning

confidence: 99%

Joint Learning of Correlation-Constrained Fuzzy Clustering and Discriminative Non-Negative Representation for Hyperspectral Band Selection

Wang

2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Spectral indices are linear or nonlinear combinations of different sensitive bands, closely related to the reflection, absorption, and growth of different plants in different spectral bands. Constructing a prediction model requires appropriate band combinations to enhance model accuracy [16]. When plants are subjected to disease stress, chlorophyll digestion, water content reduction and coverage reduction often accompany plant growth [17], leading to the degree of reflection of canopy spectral information on plant physiological growth indicators to decrease significantly [18].…”

Section: Introductionmentioning

confidence: 99%

Monitoring of Chlorophyll Content of Potato in Northern Shaanxi Based on Different Spectral Parameters

Shi,

Lu,

Sun

et al. 2024

Plants

View full text Add to dashboard Cite

Leaf chlorophyll content (LCC) is an important physiological index to evaluate the photosynthetic capacity and growth health of crops. In this investigation, the focus was placed on the chlorophyll content per unit of leaf area (LCCA) and the chlorophyll content per unit of fresh weight (LCCW) during the tuber formation phase of potatoes in Northern Shaanxi. Ground-based hyperspectral data were acquired for this purpose to formulate the vegetation index. The correlation coefficient method was used to obtain the “trilateral” parameters with the best correlation between potato LCCA and LCCW, empirical vegetation index, any two-band vegetation index constructed after 0–2 fractional differential transformation (step size 0.5), and the parameters with the highest correlation among the three spectral parameters, which were divided into four combinations as model inputs. The prediction models of potato LCCA and LCCW were constructed using the support vector machine (SVM), random forest (RF) and back propagation neural network (BPNN) algorithms. The results showed that, compared with the “trilateral” parameter and the empirical vegetation index, the spectral index constructed by the hyperspectral reflectance after differential transformation had a stronger correlation with potato LCCA and LCCW. Compared with no treatment, the correlation between spectral index and potato LCC and the prediction accuracy of the model showed a trend of decreasing after initial growth with the increase in differential order. The highest correlation index after 0–2 order differential treatment is DI, and the maximum correlation coefficients are 0.787, 0.798, 0.792, 0.788 and 0.756, respectively. The maximum value of the spectral index correlation coefficient after each order differential treatment corresponds to the red edge or near-infrared band. A comprehensive comparison shows that in the LCCA and LCCW estimation models, the RF model has the highest accuracy when combination 3 is used as the input variable. Therefore, it is more recommended to use the LCCA to estimate the chlorophyll content of crop leaves in the agricultural practices of the potato industry. The results of this study can enhance the scientific understanding and accurate simulation of potato canopy spectral information, provide a theoretical basis for the remote sensing inversion of crop growth, and promote the development of modern precision agriculture.

show abstract

“…Therefore, how to reduce the interference of redundant data and improve the efficiency of data processing is one of the key issues facing HSIs in practical applications. Feature extraction (FE) [7–9] and band selection (BS) [10–12] are the main techniques to reduce the redundancy of HSI. FE generates a discriminative representation by transforming a high‐dimensional feature space into a low‐dimensional feature space.…”

Section: Introductionmentioning

confidence: 99%

Anomaly‐background separation and particle swarm optimization based band selection for hyperspectral anomaly detection

Shang,

Duan,

Wang

et al. 2024

IET Image Processing

View full text Add to dashboard Cite

As one of the dimensionality reduction techniques of hyperspectral image (HSI), band selection (BS) does not change the spectral characteristics and physical meaning of HSIs, which is beneficial to the identification and analysis of surface objects. Recently, many BS methods for target detection have achieved promising results by making full use of the priori spectral features of the target to be detected. Conversely, anomaly detection separates the anomaly based solely on the statistical distribution difference between anomaly and background without any prior information. Therefore, the development of BS for anomaly detection has lagged far behind that of BS for target detection. To this end, this paper proposes a novel BS algorithm dedicated to anomaly detection tasks, named anomaly‐background separation and particle swarm optimization (PSO)‐based BS. Specifically, an anomaly‐background separation framework (ABSF) is established to predetermine a priori knowledge of anomaly distribution. Then, three band prioritization criteria are constructed with the anomaly‐background constraints generated by ABSF. Finally, PSO is used to find the optimal subset of bands in the solution space. The experiments on two real datasets demonstrate that the proposed method yields better detection results and greater stability compared to other BS methods discussed in this paper.

show abstract

Hyperspectral Band Selection via Optimal Combination Strategy

Cited by 11 publications

References 44 publications

Joint Learning of Correlation-Constrained Fuzzy Clustering and Discriminative Non-Negative Representation for Hyperspectral Band Selection

Joint Learning of Correlation-Constrained Fuzzy Clustering and Discriminative Non-Negative Representation for Hyperspectral Band Selection

Monitoring of Chlorophyll Content of Potato in Northern Shaanxi Based on Different Spectral Parameters

Anomaly‐background separation and particle swarm optimization based band selection for hyperspectral anomaly detection

Contact Info

Product

Resources

About