Target-adaptive CNN-based pansharpening

Abstract

We recently proposed a convolutional neural network (CNN) for remote sensing image pansharpening obtaining a significant performance gain over the state of the art. In this paper, we explore a number of architectural and training variations to this baseline, achieving further performance gains with a lightweight network which trains very fast. Leveraging on this latter property, we propose a target-adaptive usage modality which ensures a very good performance also in the presence of a mismatch w.r.t. the training set, and even across different sensors. The proposed method, published online as an off-the-shelf software tool, allows users to perform fast and high-quality CNN-based pansharpening of their own target images on general-purpose hardware.

References in corpus (4)

• Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
• Boosting the accuracy of multi-spectral image pan-sharpening by learning a deep residual network
• Land Cover Classification via Multi-temporal Spatial Data by Recurrent Neural Networks
• A Survey of Pansharpening Methods with A New Band-Decoupled Variational Model

Cited by in corpus (12)

• Hybrid Noise Removal in Hyperspectral Imagery With a Spatial-Spectral Gradient Network
• Model Inspired Autoencoder for Unsupervised Hyperspectral Image Super-Resolution
• Rethinking CNN-Based Pansharpening: Guided Colorization of Panchromatic Images via GANs
• Pansharpening by convolutional neural networks in the full resolution framework
• Unsupervised Pansharpening Based on Self-Attention Mechanism
• Full-resolution quality assessment for pansharpening
• Hyperspectral Pansharpening: Critical Review, Tools and Future Perspectives
• A New Ratio Image Based CNN Algorithm For SAR Despeckling
• Pansharpening via Frequency-Aware Fusion Network with Explicit Similarity Constraints
• S3: A Spectral-Spatial Structure Loss for Pan-Sharpening Networks
• A full-resolution training framework for Sentinel-2 image fusion
• Zero-Shot Hyperspectral Pansharpening Using Hysteresis-Based Tuning for Spectral Quality Control