Deep Learning-Based Compression Artifact Removal and Super Resolution for Aerial Imagery

FlowT-SR: A Novel Remote Sensing Image Super-Resolution Framework with Cloud Haze and Noise Suppression

A novel SR framework based on the flow matching paradigm and a diffusion transformer, named FlowT-SR, which achieves superior and reliable reconstruction quality by jointly mitigating sensor noise and thin cloud interference, achieving superior reconstruction performance compared with current state-of-the-art methods in terms of both PSNR and SSIM.

Detection-Oriented Super-Resolution for Cost-Efficient Aerial Monitoring

Spatial-channel dual autoregressive model for remote sensing image compression

Remote sensing images are typically large in size and contain abundant ground object information as well as complex spatial texture structures. These characteristics result in high storage and transmission costs, which necessitates highquality image compression methods. When compressing remote sensing images, compression models must preserve critical object boundaries and texture details under limited bandwidth while reducing the bit rate as much as possible, which poses significant challenges to the contextual modeling capability of entropy models. Although deep learningbased image compression methods have developed rapidly in recent years, they still do not fully exploit spatial-domain information. To address this issue, this paper proposes a spatial-channel dual autoregressive context modeling strategy for remote sensing images, termed Hierarchy Spatial Image Compression (HSIC). By leveraging the advantages of the window-based attention mechanism in Swin Transformer, the proposed method maintains channel-wise autoregressive modeling while progressively utilizing contextual information in the spatial domain through hierarchical modeling. This enables the model to capture potential spatial redundancy from coarse to fine granularity, thus significantly reducing bit rates without sacrificing reconstruction quality. Experimental results demonstrate that, compared with traditional image coding methods and existing learning-based image compression approaches, HSIC achieves superior rate-distortion performance across multiple remote sensing image datasets. Compared with the VTM encoder, HSIC reduces BD-rate by approximately 13.90% and 13.5% on the AID and NWPU VHR-10 datasets, respectively, and by 13.2% and 14.56% on the WorldView-3 Multispectral and Panchromatic datasets, respectively.

An end-to-end deep learning image compression method for satellite images based on entropy model

With the extensive applications of satellite image data in environmental monitoring and geographic surveying and mapping, the amount of data has increased rapidly, which brings great challenges for transmitting and storing these images. However, when processing high-resolution and multi-spectral satellite data, existing image compression methods often lead to low compression efficiency or poor reconstruction image quality due to its insufficient generalization ability. In this paper, we propose an end-to-end deep learning image compression framework for visible infrared imaging radiometer suite (VIIRS) satellite imagery. The framework consists of an analysis transform encoder, a synthesis transform decoder, a hybrid training-testing quantizer and a probability model for entropy coding. A cumulative distribution function (CDF) is constructed to compute the discrete likelihood of quantized latent symbols under a Gaussian-mixture entropy model. It is integrated with a checkerboard context structure and a VIIRS-oriented block-processing pipeline. Finally, we conduct systematic experiments based on NASA VIIRS multi-spectral datasets. Experimental results show that the proposed method achieves 0.51 ± 0.04 bpp, 38.39 ± 0.72 dB PSNR and 0.973 ± 0.007 SSIM. Relative to ELIC and the Transformer-CNN baseline, it reduced bpp by 13.6% and 10.5% and improved PSNR by 0.78 dB and 0.61 dB, respectively. The framework can compress the data volume to approximately 1.5%−4% of the original size, corresponding to an average compression ratio of about 30:1. In order to meet the processing requirements of high-resolution satellite images, we further propose a block compression strategy, which divides large-size images into sub-blocks of 256 × 256 pixels for independent compression, and realizes complete image reconstruction through decompression and splicing technology.

Satellite imagery super-resolution using GANs and aerial images

Satellite imagery often suffers from limited spatial resolution and, in many cases, high acquisition costs. These factors restrict their use in applications such as urban monitoring, land management, and wildlife studies. This work proposes an AI-based super-resolution approach that leverages high resolution aerial imagery to train a Generative Adversarial Network. Specifically, the ESRGAN (Enhanced Super-Resolution Generative Adversarial Network) architecture is adapted and trained using aerial orthophotos, enabling the transfer of learned spatial representations to low-resolution satellite images. The trained model is evaluated on satellite image patches at 2 and 4 super-resolution scales. Performance is assessed using structural, perceptual, and chromatic metrics, including SSIMY, MS-SSIM, LPIPS and CIEDE2000. The results show clear improvements, with increased sharpness, enhanced edge definition, and consistent reconstruction of urban structures and terrain features. From a quantitative perspective, the 2 scale achieves the best overall metric values, while the 4 scale maintains stable and meaningful performance despite the higher reconstruction difficulty. These findings demonstrate the feasibility of transferring super-resolution capabilities from aerial images to satellite imagery, even in the presence of spectral and geometric differences between acquisition domains. Overall, this study provides a solid foundation for the development of low-cost, AI-driven satellite image super-resolution models and outlines future research directions focused on dataset expansion, domain adaptation strategies, and sensor-specific architectural improvements.

UWResENET: A framework for Restoration and Enhancement of Underwater Imagery using Spatial Pyramid Pooling with Low-Rank Adaptation