The resolution of remote sensing images has increased with the maturation of satellite technology. Ship detection technology based on remote sensing images makes it possible to monitor a large range and far sea area, which can greatly enrich the monitoring means of maritime departments. In this paper, we conducted research on small target detection and resistance to complex background interference. First, a ship dataset with four types of targets (aircraft carriers, warships, merchant ships and submarines) is constructed, and experiments are conducted on the dataset using the object detection algorithm YOLOv4. The Kmeans++ clustering algorithm is used for a priori frame selection, and the migration learning method is used to enhance the detection effect of the YOLOv4. Second, the model is improved to address the problems of missed detection of small ships and difficulty in resisting background interference: the RFB_s (Receptive Field Block) with dilated convolution is introduced instead of the SPP (Spatial Pyramid Pooling) to enlarge the receptive field and improve the detection of small targets; the attention mechanism CBAM (Convolutional Block Attention Module) is added to adjust the weights of different features to highlight salient features useful for ship detection task, which improve the detection performance of small ships and improve the model’s ability to resist complex background. Compared to YOLOv4, our proposed model achieved a large improvement in mAP (mean Average Precision) from 77.66% to 91.40%.

The detection of vehicle targets in infrared aerial remote sensing images captured by drones presents challenges due to a significant imbalance in vehicle distribution, complex backgrounds, the large scale of vehicles, and the dense and arbitrarily oriented distribution of targets. The RYOLOv5_D model is proposed based on the YOLOv5-obb rotation model. Firstly, we reconstruct a new vehicle remote sensing dataset, BalancedVehicle, to achieve data balance. Secondly, given the challenges of complex backgrounds in infrared remote sensing images, the AAHE method is proposed to highlight infrared remote sensing vehicle targets while reducing background interference during the detection process. Moreover, in order to address the issue of detecting challenges under complex backgrounds, the CPSAB attention mechanism is proposed, which could be used together with DCNv2. GSConv is also used to reduce the model parameters while ensuring accuracy. This combination could improve the model’s generalization ability and, consequently, enhance the detection accuracy for various vehicle categories. The RYOLOv5s_D model, trained on the self-built dataset BalancedVehicle, demonstrates a notable improvement in its mean average precision (mAP), increasing from 73.6% to 78.5%. Specifically, the average precision (AP) for large aspect ratio vehicles such as trucks and freight cars increases by 11.4% and 8%, respectively. The RYOLOv5m_D and RYOLOv5l_D models achieve accuracies of 82.6% and 84.3%. The Param of RYOLOv5_D is similar to that of the YOLOv5-obb, while possessing a decrease in computational complexity of 0.6, 4.5, and 12.8GFLOPS. In conclusion, the RYOLOv5_D model’s superior accuracy and real-time capabilities in infrared remote sensing vehicle scenarios are validated by comparing various advanced models based on rotation boxes on the BalancedVehicle dataset.

Despite significant challenges, the text-guided remote sensing image generation method shows great potential in many practical applications such as generative adversarial networks in remote sensing tasks; generated images still face challenges such as low realism, face challenges, and unclear details. Moreover, the inherent spatial complexity of remote sensing images and the limited scale of publicly available datasets make it particularly challenging to generate high-quality remote sensing images from text descriptions. To address these challenges, this paper proposes the RSVQ-Diffusion model for remote sensing image generation, achieving high-quality text-to-remote-sensing image generation applicable for target detection, simulation, and other fields. Specifically, this paper designs a spatial position encoding mechanism to integrate the spatial information of remote sensing images during model training. Additionally, the Transformer module is improved by incorporating a short-sequence local perception mechanism into the diffusion image decoder, addressing issues of unclear details and regional distortions in generated remote sensing images. Compared with the VQ-Diffusion model, our proposed model achieves significant improvements in the Fréchet Inception Distance (FID), the Inception Score (IS), and the text–image alignment (Contrastive Language-Image Pre-training, CLIP) scores. The FID score successfully decreased from 96.68 to 90.36; the CLIP score increased from 26.92 to 27.22, and the IS increased from 7.11 to 7.24.

Vehicle detection with optical remote sensing images has become widely applied in recent years. However, the following challenges have remained unsolved during remote sensing vehicle target detection. These challenges include the dense and arbitrary angles at which vehicles are distributed and which make it difficult to detect them; the extensive model parameter (Param) that blocks real-time detection; the large differences between larger vehicles in terms of their features, which lead to a reduced detection precision; and the way in which the distribution in vehicle datasets is unbalanced and thus not conducive to training. First, this paper constructs a small dataset of vehicles, MiVehicle. This dataset includes 3000 corresponding infrared and visible image pairs, offering a more balanced distribution. In the infrared part of the dataset, the proportions of different vehicle types are as follows: cars, 48%; buses, 19%; trucks, 15%; freight, cars 10%; and vans, 8%. Second, we choose the rotated box mechanism for detection with the model and we build a new vehicle detector, ML-Det, with a novel multi-scale feature fusion triple cross-criss FPN (TCFPN), which can effectively capture the vehicle features in three different positions with an mAP improvement of 1.97%. Moreover, we propose LKC–INVO, which allows involution to couple the structure of multiple large kernel convolutions, resulting in an mAP increase of 2.86%. We also introduce a novel C2F_ContextGuided module with global context perception, which enhances the perception ability of the model in the global scope and minimizes model Params. Eventually, we propose an assemble–disperse attention module to aggregate local features so as to improve the performance. Overall, ML-Det achieved a 3.22% improvement in accuracy while keeping Params almost unchanged. In the self-built small MiVehicle dataset, we achieved 70.44% on visible images and 79.12% on infrared images with 20.1 GFLOPS, 78.8 FPS, and 7.91 M. Additionally, we trained and tested our model on the following public datasets: UAS-AOD and DOTA. ML-Det was found to be ahead of many other advanced target detection algorithms.

High-resolution optical images will provide planetary geology researchers with finer and more microscopic image data information. In order to maximize scientific output, it is necessary to further increase the resolution of acquired images, so image super-resolution (SR) reconstruction techniques have become the best choice. Aiming at the problems of large parameter quantity and high computational complexity in current deep learning-based image SR reconstruction methods, we propose a novel Recursive Swin Transformer Super-Resolution Network (RSTSRN) for SR applied to images. The RSTSRN improves upon the LapSRN, which we use as our backbone architecture. A Residual Swin Transformer Block (RSTB) is used for more efficient residual learning, which consists of stacked Swin Transformer Blocks (STBs) with a residual connection. Moreover, the idea of parameter sharing was introduced to reduce the number of parameters, and a multi-scale training strategy was designed to accelerate convergence speed. Experimental results show that the proposed RSTSRN achieves superior performance on 2×, 4× and 8×SR tasks to state-of-the-art methods with similar parameters. Especially on high-magnification SR tasks, the RSTSRN has great performance superiority. Compared to the LapSRN network, for 2×, 4× and 8× Mars image SR tasks, the RSTSRN network has increased PSNR values by 0.35 dB, 0.88 dB and 1.22 dB, and SSIM values by 0.0048, 0.0114 and 0.0311, respectively.