Instance segmentation in remote sensing images (RSI) poses significant challenges due to the diverse scales of targets, scene complexity, and a high number of targets, making most methods struggle with suboptimal performance and time-consuming computations. To solve those problems, a fast and accurate RSI instance segmentation model (named DCTC) is designed in this article. DCTC transforms classification problem into regression problem to improve the reference speed. DCTC contains two parallel branches. The contour branch performs iterative regression on contours, extracting precise contour information to improve boundary accuracy. Meanwhile, the discrete cosine transformation (DCT) branch refines mask predictions and supplements instance context information, which particularly benefits the segmentation of small targets. DCT encoding is employed in the DCT branch to convert the mask representation into DCT format, aligning the outputs of the contour and DCT branches. Three innovative modules are introduced in the DCT branch: the coarse result generation (CRG) module, iteratively deform and regression (IDR) module, and contour and DCT fusion module (CDF). The CRG module generates coarse DCT vectors and contour coordinates, facilitating information exchange between the contour and DCT branches. The IDR module iteratively refines DCT vectors, enabling DCTC to focus more on small targets and instance details. The CDF module merges DCT vectors and contour coordinates, ensuring effective interaction between boundary and context information, thereby enhancing performance. Extensive experiments demonstrate the superiority of DCTC, which achieves 67.7, 36.3, 67.4, and 55.1AP on NWPU VHR-10, iSAID, synthetic aperture radar (SAR) ship detection dataset, and high-resolution SAR images dataset, respectively, and ranks first among state-of-the-art methods while maintaining real-time processing capability. Furthermore, DCTC exhibits strong performance on both optical and SAR images, and the designed DCT branch can be simply plug into any contour-based method to improve the network performance.

In remote sensing image classification task, it is often apply a model trained on one dataset (source domain) to another dataset (target domain). However, due to the presence of domain shift between these domains where data are not independent and identically distributed, the performance of the model typically deteriorates. Domain adaptation aims to improve the generalization performance of the model in the target domain. In response to the challenges of intricate backgrounds, domain shift, and potentially unlabeled target domain in remote sensing images, this paper proposes a network specifically designed for unsupervised domain adaptation (UDA) classification of remote sensing images, named PPLM-net. The network consists of a domain adversarial training (DAT) module, a partial patch local masking (PPLM) module and a teacher-student network module. The DAT module enables the network to extract domain-invariant features. The PPLM module compels the model to focus on the global information of target domain remote sensing images with intricate backgrounds, learning contextual content to improve model performance. The teacher network generates pseudo-labels for complete unlabeled target domain images. The student network trained with PPLM target domain classification loss to generate robust and discriminative features. We construct a dataset dedicated to the UDA scene classification task of remote sensing images named RSDA. We collect images from 4 publicly available datasets spanning 7 common categories, containing over 10,000 images. Compared with the current state-of-the-art UDA model, PPLM-net achieves the best results in 12 domain adaptation classification tasks on RSDA. The average accuracy reaches 99.115%.

Instance segmentation in remote sensing imagery is a significant application area within computer vision, holding considerable value in fields such as land planning and aerospace. The target scales of remote sensing images are often small, the contours of different categories of targets can be remarkably similar, and the background information is complex, containing more noise interference. Therefore, it is essential for the network model to utilize the background and internal instance information more effectively. Considering all the above, to fully adapt to the characteristics of remote sensing images, a network named CGNet, which combines an enhanced backbone with a contour–mask branch, is proposed. This network employs gated recurrent units for the iteration of contour and mask branches and adopts the attention head for branch fusion. Additionally, to address the common issues of missed and misdetections in target detection, a supervised backbone network using contrastive pretraining for feature supplementation is introduced. The proposed method has been experimentally validated in the NWPU VHR-10 and SSDD datasets, achieving average precision metrics of 68.1% and 67.4%, respectively, which are 0.9% and 3.2% higher than those of the suboptimal methods.