EORestore-Agent: Fidelity-Guided Agentic Restoration of Remote Sensing Images with Composite Degradations

2026-10-05

AI Summary

This study addresses the challenge of tool selection in compound degradation restoration for remote sensing images, which arises from the absence of reference-free validation. To this end, we propose a traceable agent framework based on vision-language models (VLMs). Methodologically, by fine-tuning a VLM alongside a relative quality scorer and incorporating synthetic degradation chain supervision with a stepwise acceptance strategy, our approach replaces the conventional full-reference paradigm with a mechanism that progressively evaluates residual degradation and predicts quality gains, thereby enabling reference-free, interpretable, step-by-step decision-making. Experimental results demonstrate that the proposed method improves PSNR by 2.3 to 3.2 dB over the strongest baseline on synthetic benchmarks, significantly outperforming zero-shot agents designed for natural images, while exhibiting effective transferability to real-world atmospheric degradation scenarios.

Abstract

Remote sensing images often carry composite degradations, in which haze, cloud, noise, blur, low light, and low resolution coexist. Restoring them requires deciding which tool to apply, in what order, and when to stop, yet no clean reference is available at inference time to verify these decisions. All-in-one models trained on single degradations converge to a narrow PSNR band as degradations accumulate. To formulate real-world remote sensing restoration as a traceable trajectory, we present EORestore-Agent, which replaces this unmeasurable objective with reference-free, verifiable per-step decisions. A fine-tuned vision-language model reports all residual degradation types, whose tool pools are scored together, so the restoration order emerges from step-wise selection. A relative quality scorer, trained with full-reference supervision on synthetic degradation chains, predicts the changes in PSNR, SSIM, and LPIPS from the current image to each candidate. A step is accepted only when no predicted change is negative and the predicted PSNR gain is positive. Otherwise, the agent keeps the current image. On a synthetic Landsat-8 benchmark with six degradation types, EORestore-Agent improves PSNR by 2.3 to 3.2 dB over the strongest retrained all-in-one baseline on composites of two to six degradations, whereas zero-shot natural-image agents fall below the degraded input in PSNR in 17 of 18 settings. Replacing the learned scorer with no-reference quality differences costs 1.1 to 4.6 dB. The remaining harmful steps are small and cluster near the acceptance threshold. Sentinel-2 examples illustrate transfer to real atmospheric degradation without retraining.

Problem

• remote sensing image restoration
• composite degradations
• reference-free restoration
• all-in-one model
• degradation trajectory

Innovation

• Agentic Restoration
• Vision-Language Model
• Relative Quality Scorer
• Composite Degradations
• Reference-free Verification