Instance segmentation predicts the categories of all instances and locates them using pixel-level masks. Although existing methods have shown exemplary performance, the poor boundaries due to the lack of fine-grained information re-mains a challenge for RSIs instance segmentation. In this paper, to address the problem, we propose a novel instance segmentation branch, namely Mask Decoupled Head, which is mainly composed of a Feature Enhance Module (FEM) and a Feature Decoupled Module (FDM). FEM enhances the rep-resentation of the body features through the low-frequency component of images. FDM decouples the segmentation task by supervising body and edge separately and leverages fine-grained information to complement the boundary details. We performed comprehensive experiments on NWPU VHR-10 and HRSID datasets to evaluate the effectiveness of our pro-posed method and achieved good performance. © 2022 IEEE.

Road extraction faces the great challenges of occlusion, large span, and complex backgrounds in remote sensing images. Many existing methods receive context from regions near the road non-differently, and the context from irrelevant regions instead harms the semantics of features and leads to the mis-classification of the network. To address the above problem, we propose a Semantic-Aware Context Module (SACM) that encourages the network to model the context of different se-mantics supervised by a soft foreground map. And Strip Pooling Module (SPM) is introduced to match the fact that roads tend to be strip-shaped, contributing to the suppression of contamination information in irrelevant regions. Both SACM and SPM enable the network to obtain more specific semanti-cally relevant context. The experimental results on the Deep-Globe dataset show that the proposed method tremendously improves the performance of the network. © 2022 IEEE.

Hyperspectral image (HSI) classification has attracted ever-rising attention to better performance based on limited labeled data. In this paper, a domain adaptation method of class balance based on few-shot learning is proposed, which obtains the classification results of target HSI by training the dataset in the source domain containing sufficient labeled data. We use a random weighted sampling strategy in the source domain and the generative adversarial network (GAN) in the target domain to reduce the label distribution shift caused by unbalanced classes. Then, the conditional maximum mean discrepancy (CMMD) is presented for a more comprehensive domain alignment by considering the posterior data distribution. In addition, the double cross non-local block and multi-scale strategy are adopted in the feature extraction stage to get a refined classification result. Experimental results on public HSI datasets demonstrate that our method is efficient and outperforms other baselines. © 2022 IEEE.

With the development of remote sensing satellite technology, the resolution of remote sensing images is constantly improved, but there are difficulties in obtaining labeled SAR image datasets for target detection and recognition. To address the problem that only limited SAR image target detection and recognition data are available, a novel data augmentation method based on convolutional neural network is proposed. Firstly, the Synthetic Aperture Radar (SAR) image target detection and recognition dataset SAR_OD was produced based on the synthesis of military targets and background images in MSTAR dataset. But considering the fact that the number of targets in each image in SAR_OD is still not enough for training a target detection model with good performance, we augmented SAR_OD and then we obtained SAR_OD+ dataset. It is proved that the model trained on SAR_OD+ dataset is significantly improved in the evaluation index by the data augmentation method proposed in this paper, especially in the experiments using only 50% of the training data. Therefore, the proposed data augmentation method can be used to improve the performance of SAR image target detection and recognition model in the case of limited labeled data. © 2021 IEEE.

Multi-class geospatial object detection in remote sensing images suffer great challenges, such as large scales variability and complex background. Although feature pyramid network (FPN) can alleviate the problem of scale variation to some extent, it causes the loss of spatial and semantic information which is not conducive to object location. To address the above problem, this paper proposes a discriminative feature pyramid network (DFPN) by introducing a global guidance module (GGM) and a feature aggregation module (FAM). Specifically, the global guidance module delivers the high-level semantic information to lower layers, so as to obtain feature maps with stronger semantic information to eliminate the interference caused by complex background. The feature aggregation module enhances the interflow of information between different layers and better captures the discrimination information at each layer. We validate the effectiveness of our method on the NWPU VHR-10 and RSOD datasets, the results outperform baseline by 2.06 and 3.88 points respectively. © 2020 IEEE.

Remote sensing images play an important role in various applications. To make it easier for humans to understand remote sensing images, the task of remote sensing image captioning attracts more and more researchers' attention. Inspired from the way human receives visual information, attention mechanism has been widely used in remote sensing image understanding. To catch more scene information and improve the stability of the generated sentences, a new attention mechanism called scene attention is proposed. Except for the current attention via the current hidden state of the long shortterm memory network (LSTM), our proposed method simultaneously explores the global visual information from the mean feature of all convolutional features. The effectiveness of the proposed method is evaluated on UCM-captions, Sydney-captions and RSICD datasets. The results of our experiment show that comparing with some other captioning methods, our method is more stable and obtains a better performance. © 2020 IEEE.

Synthetic aperture radar (SAR) image despeckling is recognized as the basis for SAR image processing and interpretation. Over the past decades, many impressed speckle reduction methods have been developed and achieved good performance under certain circumstances. However, how to suppress speckle noise in a homogeneous region while more effectively protecting details and avoid distortion of data features caused by homomorphic transformation is still an urgent problem. In this paper, a novel speckle reduction algorithm based on generative adversarial network (GAN) is proposed, which contains a generator and a discriminator. For the generator that is used directly for subsequent noise reduction, a total variation (TV) loss function is added. Meanwhile, we directly learn the mapping between the input image and the ground truth rather than the logarithmic transformation. Indeed, the improved lightweight discriminative network will also provide learning guidance for the generator. Experiments on simulatedSAR images and real SAR images demonstrate the improvement in visual and statistical performance comparing to the state-of-the-art despeckling algorithms. © 2020 IEEE.