Hyperspectral image (HSI) classification accuracy has been greatly improved by employing deep learning. The current research mainly focuses on how to build a deep network to improve the accuracy. However, these networks tend to be more complex and have more parameters, which makes the model difficult to train and easy to overfit. Therefore, we present a lightweight deep convolutional neural network (CNN) model called S2FEF-CNN. In this model, three S2FEF blocks are used for the joint spectral–spatial features extraction. Each S2FEF block uses 1D spectral convolution to extract spectral features and 2D spatial convolution to extract spatial features, respectively, and then fuses spectral and spatial features by multiplication. Instead of using the full connected layer, two pooling layers follow three blocks for dimension reduction, which further reduces the training parameters. We compared our method with some state-of-the-art HSI classification methods based on deep network on three commonly used hyperspectral datasets. The results show that our network can achieve a comparable classification accuracy with significantly reduced parameters compared to the above deep networks, which reflects its potential advantages in HSI classification.

Hyperspectral remote sensing has been widely used in various fields. The classification of different feature categories is one of the main research directions. For different application scenarios, considering the interference of different environments, it brings great challenges to the recognition of ground feature classification. In this paper, hyperspectral remote sensing data sets under different scenarios are introduced into the classifier for classification and recognition, The rich spectral information features in hyperspectral data are used for hard classification, and the classification results are used as the discrimination basis for the classification challenges of different data sets, and the recognition results are summarized.

In recent years, researches on combining wavelet decomposition and convolutional neural network (CNN) together to classify hyperspectral images (HSI) have emerged, and some effective classification models have been proposed and achieved good classification results. However, there are two problems for most of the proposed models. One is the heavy training parameters and the other is that there is no distinction between the classification effectiveness of low frequency and high frequency features after wavelet decomposition. In this paper, a new light-weighted HSI classification model (LLFWCNN) is proposed, which performs multi-layer wavelet decomposition for HSI after dimensionality reduction, and only arranges the low frequency features in a specific stack mode, then classifies them through a well-designed convolutional neural network. Compared with other classification models, the number of parameters is only 2.5% of that of other models. And only the low frequency features after wavelet decomposition are used, while the high frequency features are abandoned. The results showed that compared with the state-of-the-art classification models, LLFWCNN could obtain the same or even better classification results with fewer network parameters and proved that the low frequency features after wavelet decomposition provide LLFWCNN with more favorable information for HSI classification as well.

Supervised classification is one of the widespread applications in hyperspectral data analysis. Due to the large number of hyperspectral data bands and the redundancy of information between the bands, it brings great challenges to hyperspectral classification. The effect of hyperspectral data feature extraction determines the performance of classification accuracy. In order to improve the classification accuracy, this paper proposes a joint feature extraction method based on minimum noise fraction (MNF) and variational mode decomposition (VMD). The hyperspectral data is firstly minimum noise fraction transformed, and then the first few MNF sequences containing the main information of the hyperspectral are subjected to VMD. Then, identify and classify each mode component obtained by decomposing. Finally, through the support vector machine (SVM) classification and comparative analysis, the method has a good accuracy.