Abstract
To improve the accuracy and generalization ability of hyperspectral image classification, a feature extraction method integrating principal component analysis (PCA) and local binary pattern (LBP) is developed for hyperspectral images in this article. The PCA is employed to reduce the dimension of the spectral features of hyperspectral images. The LBP with low computational complexity is used to extract the local spatial texture features of hyperspectral images to construct multifeature vectors. Then, the gray wolf optimization algorithm with global search capability is employed to optimize the parameters of kernel extreme learning machine (KELM) to construct an optimized KELM model, which is used to effectively realize a hyperspectral image classification (PLG-KELM) method. Finally, the Indian pines dataset, Houston dataset, and Pavia University dataset and an application of WHU-Hi-LongKou dataset are selected to verify the effectiveness of the PLG-KELM. The comparison experiment results show that the PLG-KELM can obtain higher classification accuracy, and takes on better generalization ability for small samples. It provides a new idea for processing hyperspectral images.
Highlights
In order to improve the classification accuracy, the hyperspectral image features based on spatial spectrum information is extracted and a PLG-kernel extreme learning machine (KELM) method based on PCA, local binary pattern (LBP), GWO and KELM is developed for hyperspectral image classification
The classification models of support vector machine (SVM), broad learning system (BLS), contractive auto-encoder and convolutional neural network(CAE-CNN), PCA-CNN and KELM are selected to compare with the PLG-KELM[57]
For the PLG-KELM, it can be seen that the overall accuracy (OA), AA and Kappa coefficient of the 0.5% training set are improved by 0.06%, 0.5% and 0.08% than those of 1% training set in small samples, which indicate that the PLG-KELM has better classification accuracy and generalization ability
Summary
THE hyperspectral sensor can obtain approximately continuous spectral curves of ground and objects in a large number of electromagnetic bands, such as ultraviolet, visible, near-infrared and mid-infrared by combining the spectral information of the reflection features of ground and objects. Mou et al.[32] proposed a novel graph-based semisupervised network called nonlocal graph convolutional network These propsoed methods can better realize image classification and obtain better classification accuracy, but they still exist some shortcomings, such as poor generalization ability, higher complexity, lower accuracy for small samples and so on. Chen et al.[44] proposed a KELM classification algorithm with space spectrum information, which fused spectral information and spatial information in the construction of hyperspectral image feature vector to improve the classification accuracy of KELM. In order to improve the accuracy and efficiency of hyperspectral image classification, the PCA, LBP, GWO and KELM are integrated to develop a new hyperspectral image classification method, namely PLG-KELM in this paper.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
