Fine-grained land cover classification from high-resolution remote sensing imagery plays a vital role in urban and environmental monitoring. While existing spatial-domain based approaches achieve notable progress, their performance in complex scenarios remains constrained by insufficient modeling of characteristics. This letter proposes the continuous low-frequency decomposition architecture (CLFDA) to address insufficient cross-domain modeling of multi-scale frequency characteristics in current methods. The architecture introduces frequency domain features through continuous low-frequency decomposition, where each frequency decomposition and enhancement module employ discrete wavelet transform to separate spatial and low-frequency features into low-frequency and high-frequency subbands. Low-frequency features feed back into the encoder for global context, while high-frequency features are routed to the decoder via attention mechanisms for detail refinement, enabling bidirectional spatial-frequency fusion. By integrating convolutional neural networks, vision transformer, and mamba backbones, our CLFDA achieves 2.0% and 3.46% averagemIoUimprovements on the GID-15 and the FUSU datasets, respectively. These consistent performance gains across heterogeneous backbones demonstrate the effectiveness and generalizability of our CLFDA in modeling frequency domain features. The code is at https://github.com/GeoRSAI/CLFDA.

Domain adaptation is a pivotal technique for improving the classification performance of remote sensing scenes impacted by data distribution shifts. Existing spatial-domain feature alignment methods are vulnerable to complex scene clutter and spectral variations. Considering the robustness of frequency representation in preserving edge details and structural patterns, this paper presents a novel spatial-frequency multiple alignment domain adaptation (SFMDA) method for remote sensing scene classification. First, a frequency-domain invariant feature learning module is introduced, which employs the Fourier transform and high-frequency mask strategy to derive frequency-domain features exhibiting enhanced inter-domain invariance. Subsequently, a spatial-frequency feature cross fusion module is developed to achieve more robust and domain-representative spatial-frequency fusion representations through dot product attention and interaction mechanisms. Finally, a multiple feature alignment strategy is devised to minimize both spatial-domain feature differences and fusion feature discrepancies across the source and target domains, thereby facilitating more effective inter-domain knowledge transfer. Experimental results on six cross-domain scenarios demonstrate that SFMDA outperforms eight state-of-the-art methods, achieving a 3.87%–17.98% accuracy improvement. Furthermore, SFMDA is compatible with existing spatial-domain learning frameworks, enabling seamless integration for further performance gains. Our code will be available at https://github.com/GeoRSAI/SFMDA.

Benefitting from the easy availability of image-level labels, weakly supervised semantic segmentation (WSSS) methods based on class activation maps (CAMs) have made significant progress in building extraction from remote sensing imagery. However, image-level labels lack precise spatial locations and boundary ranges of buildings, posing challenges in achieving comprehensive and structurally clear building extraction. Furthermore, due to the complex background interference and the diversity of building in high-resolution remote sensing imagery (HRRS), small and sparse buildings suffer from insufficient attention in CAMs. To solve the above problems, this article proposes a spatial structure constraint and local semantic awareness-based WSSS method, CANet, for extracting buildings from HRRS. Specifically, we design a spatial structure constraint module to generate CAMs with finer spatial structural details of buildings, which minimizes feature differences from patches of different granularities and the whole image. Moreover, a local semantic awareness module is designed to address the issue of insufficient coverage of CAMs on sparse and tiny building. This module first strengthens the feature extraction network by embedding discriminative suppression units to force the network to focus on more nondiscriminative regions. Subsequently, visual word learning is introduced to identify additional object categories. Finally, four WSSS datasets are constructed based on public datasets with two simple and two complex scenarios. The results demonstrate that the proposed method outperforms 11 state-of-the-art methods, improving the intersection over union by at least 3.46% and 0.29% in both simple and complex scenarios, respectively.

Integrated Sentinel-1 synthetic aperture radar (SAR) imagery and Sentinel-2 optical imagery have shown great promise in mapping large-scale building height. Effectively fusing the complementary features of SAR and optical imagery is a key challenge in enhancing the building height estimation performance. However, SAR imagery and optical imagery have significant heterogeneity, which makes obtaining accurate building height a challenging problem. In this article, we propose a hybrid multimodal fusion network (MF-BHNet) for building height estimation using Sentinel-1 SAR imagery and Sentinel-2 optical imagery. First, we design a hybrid multimodal encoder to mine modal-specific feature and model intermodal correlation. In particular, an intramodal encoder (IME) is designed to reconstruct valuable intramodal information, and a transformer-based cross-modal encoder (CME) is used to model intermodal correlation and capture contextual information. Then, a coarse-fine progressive multimodal fusion method is proposed to fuse SAR feature and optical feature to improve the building height estimation performance. We construct a building height dataset by introducing superior building footprints to validate our method. Experimental results demonstrate that our MF-BHNet method outperforms the compared 11 state-of-the-art methods, which achieves the lowest root-mean-square error (RMSE) of 3.6421 m. Besides, compared to the four publicly available building height products, the mapping result of the proposed method has significant advantages in terms of spatial detail and accuracy.

Recent advances in deep learning have dramatically improved the performance of content-based remote sensing image retrieval (CBRSIR) with the same distribution of training set (source domain) and test set (target domain). In fact, their distributions are inconsistent in most cases, which can lead to a dramatic decrease in retrieval performance. Currently, some unsupervised domain adaptation (DA) methods for other remote sensing applications have been proposed to eliminate the inconsistency. However, the current unsupervised DA methods do not make full use of the target domain’s distribution characteristics when delineating its decision boundary. This tends to degrade the cross-domain retrieval performance. In this article, a pseudo-label consistency learning-based unsupervised DA method (PCLUDA) is proposed for cross-domain CBRSIR. Our PCLUDA method minimizes the difference in probability distribution between the target domain and its perturbed output by a pseudo-label self-training and consistency regularization strategy, followed by adjusting the target domain’s decision boundaries to the low-density region. Besides, minimize class confusion (MCC) is introduced to reduce negative transfer caused by large intraclass variance of RSIs. Two cross-domain datasets with 12 cross-domain scenarios are constructed based on six open access datasets to measure DA methods. Experimental results show that our PCLUDA method achieves superior retrieval performances with average retrieval precision improvement by 4.9%–32.3% compared with eight state-of-the-art DA approaches in complex cross-domain scenarios. Furthermore, other experimental results indicate that our PCLUDA can also reach optimal retrieval performances in different kinds of deep learning networks [i.e., vision transformer (ViT) and convolutional neural networks (CNNs)].

Unsupervised domain adaptation (UDA) is an important solution to reduce the bias between the labeled source domain and the unlabeled target domain. It has attracted more attention for optical remote sensing image scene classification and retrieval. Currently, most of the previous work is devoted to closed-set UDA. In fact, the target domain often contains unknown classes. Moreover, some open UDA methods mine structural information of the target domain directly from the type knowledge of the source domain, and less directly from the unlabeled data of the target domain. In this paper, we propose a new self-supervised-driven open-set UDA method combining contrastive self-supervised learning with consistency self-training for optical remote sensing scene classification and retrieval. Specifically, a contrastive self-supervised learning network is introduced to learn discriminative features from the unlabeled target domain data. Moreover, a novel open-set class learning module is developed based on two-level confidence rules and the consistency self-training strategy, which can obtain reliable unknown class samples for co-training. Finally, an open-set dataset including six cross-domain scenarios is constructed based on three public datasets and several experiments are conducted with eleven state-of-the-art domain adaptation methods. Experimental results demonstrate that our proposed method achieves superior performances on the six open-set cross-domain scenarios in both scene classification and retrieval. Especially, our method improves the overall classification accuracies by 9.72% to 24.06% and improve mean average retrieval precisions by 8.06% to 16.21% on the complex UCMD (source domain) → NWPU (target domain) scenario, compared with the other eleven state-of-the-art methods.

The feature representation capability of the object detection model is considerably reduced if the available training samples are few-shot. The challenges of arbitrary orientation and complex background of ground objects are universal in very high spatial resolution remote sensing imageries, resulting in massive difficulty on few-shot object detection task. However, existing methods for few-shot object detection are not explored in terms of the capabilities of feature representation in remote sensing images. To solve these issues, we propose a few-shot object detection method incorporating a multiscale object contrastive learning. First, our method performs contrastive learning to represent the object feature fully by adopting Siamese network structure in the few-shot training. On the one hand, the Siamese structure’s lower branch is embedded in the contrastive learning process to cope with the challenge of the complexity of images; on the other hand, a multiscale instance feature module is designed to obtain multiscale contrastive information. Second, we leverage the proposed contrastive multiscale proposal loss (CMSP Loss) to make full use of multiscale information. It can promote our method to fit the data better. The experimental results show that the proposed method has good feature representation capabilities in the few-shot object detection. Moreover, the performance of the proposed method is better than that of related methods.

The detection accuracy of the grounded-wire transient electromagnetic method is affected by the incorrect geoelectric structure and burial depth due to shadow and source overprint effects. To better understand these effects, the three conditions under which these effects are present are analyzed separately. First, the analytical expressions are derived for computing the response of the formative wave and the surface wave in homogenous earth. The effects of the formative wave at the anomaly's location and at the receiving site are analyzed. Then, the response variation due to the presence of an anomaly between the transmitter and the receiver is quantitatively analyzed using the finite-difference method. Finally, the effects of shadow and source overprint on the data inversion are investigated, which is verified by a case study. The wrong geoelectric information is caused by the shadow and source overprint effects. To avoid this, multisource observation can be used to weaken and utilize the shadow and source overprint effects.