Imaging spectroscopy, also known as hyperspectral imaging, has been transformed in the last four decades from being a sparse research tool into a commodity product available to a broad user community. Specially, in the last 10 years, a large number of new techniques able to take into account the special properties of hyperspectral data have been introduced for hyperspectral data processing, where hyperspectral image classification, as one of the most active topics, has drawn massive attentions. Spectral-spatial hyperspectral image classification can achieve better classification performance than its pixel-wise counterpart, since the former utilizes not only the information of spectral signature but also that from spatial domain. In this paper, we provide a comprehensive overview on the methods belonging to the category of spectral-spatial classification in a relatively unified context. First, we develop a concept of spatial dependency system that involves pixel dependency and label dependency, with two main factors: neighborhood covering and neighborhood importance. In terms of the way that the neighborhood information is used, the spatial dependency systems can be classified into fixed, adaptive, and global systems, which can accommodate various kinds of existing spectral-spatial methods. Based on such, the categorizations of single-dependency, bilayer-dependency, and multiple-dependency systems are further introduced. Second, we categorize the performings of existing spectral-spatial methods into four paradigms according to the different fusion stages wherein spatial information takes effect, i.e., preprocessing-based, integrated, postprocessing-based, and hybrid classifications. Then, typical methodologies are outlined. Finally, several representative spectral-spatial classification methods are applied on real-world hyperspectral data in our experiments.

Self-supervised pretraining bears the potential to generate expressive representations from large-scale Earth observation (EO) data without human annotation. However, most existing pretraining in the field is based on ImageNet or medium-sized, labeled remote sensing (RS) datasets. In this article, we share an unlabeled dataset Self-Supervised Learning for Earth Observation-Sentinel-1/2 ( SSL4EO - S12 ) to assemble a large-scale, global, multimodal, and multiseasonal corpus of satellite imagery. We demonstrate SSL4EO-S12 to succeed in self-supervised pretraining for a set of representative methods: momentum contrast (MoCo), self-distillation with no labels (DINO), masked autoencoders (MAE), and data2vec, and multiple downstream applications, including scene classification, semantic segmentation, and change detection. Our benchmark results prove the effectiveness of SSL4EO-S12 compared to existing datasets. The dataset, related source code, and pretrained models are available at https://github.com/zhu-xlab/SSL4EO-S12 .

Recently, many convolutional neural network (CNN) methods have been designed for hyperspectral image (HSI) classification since CNNs are able to produce good representations of data, which greatly benefits from a huge number of parameters. However, solving such a high-dimensional optimization problem often requires a large number of training samples in order to avoid overfitting. In addition, it is a typical nonconvex problem affected by many local minima and flat regions. To address these problems, in this article, we introduce the naive Gabor networks or Gabor-Nets that, for the first time in the literature, design and learn CNN kernels strictly in the form of Gabor filters, aiming to reduce the number of involved parameters and constrain the solution space and, hence, improve the performances of CNNs. Specifically, we develop an innovative phase-induced Gabor kernel, which is trickily designed to perform the Gabor feature learning via a linear combination of local low-frequency and high-frequency components of data controlled by the kernel phase. With the phase-induced Gabor kernel, the proposed Gabor-Nets gains the ability to automatically adapt to the local harmonic characteristics of the HSI data and, thus, yields more representative harmonic features. Also, this kernel can fulfill the traditional complex-valued Gabor filtering in a real-valued manner, hence making Gabor-Nets easily perform in a usual CNN thread. We evaluated our newly developed Gabor-Nets on three well-known HSIs, suggesting that our proposed Gabor-Nets can significantly improve the performance of CNNs, particularly with a small training set.

Active learning (AL) has obtained a great success in supervised remotely sensed hyperspectral image classification, since it is able to select highly informative training samples. As an intrinsically biased sampling approach, AL generally favors the selection of samples following discriminative distributions, which are located in low-density areas. However, hyperspectral data are often highly class-mixed, i.e., most samples fluctuate in the overlapping regions of distributions of different classes. In this case, the potential of AL to select effective training samples is more limited. As AL strongly depends on the features, a possibility to increase its capabilities is to transfer the data into a highly discriminative feature space, in which the mixture of distributions that different classes of data follow tends to reduce. Based on this observation, in this paper, we introduce the concept of feature-driven AL, namely, the sample selection is going to be conducted in a given optimized feature space whose superiority is measured by an overall error probability. For illustrative purposes, we used Gabor filtering and morphological profiles for instantiation. Our experimental results, obtained on three real hyperspectral data sets, indicate that the proposed approach can significantly improve the potential of AL for hyperspectral image classification.

In this work, we propose a new semisupervised active learning approach for hyperspectral image classification. The proposed method aims at improving machine generalization by using pseudolabeled samples, both confident and informative, which are automatically and actively selected, via semisupervised learning. The learning is performed under two assumptions: a local one for the labeling via a superpixel-based constraint dedicated to the spatial homogeneity and adaptivity into the pseudolabels, and a global one modeling the data density by a multinomial logistic regressor with a Markov random field regularizer. Furthermore, we propose a density-peak-based augmentation strategy for pseudolabels, due to the fact that the samples without manual labels in their superpixel neighborhoods are out of reach for the automatic sampling. Three real hyperspectral datasets were used in our experiments to evaluate the effectiveness of the proposed superpixel-based semisupervised learning approach. The obtained results indicate that the proposed approach can greatly improve the potential for semisupervised learning in hyperspectral image classification.