AnyTSR: Any-Scale Thermal Super-Resolution for UAV

2025-04-18

AI Summary

Thermal imaging sensors on UAVs suffer from inherently low resolution, leading to loss of fine details and blurred boundaries; existing super-resolution (SR) methods are typically limited to fixed scaling factors, computationally expensive, and lack deployment flexibility. To address this, we propose the first single-model thermal image SR method supporting arbitrary continuous scaling factors. Our approach introduces a novel upsampling module that integrates an explicit feature encoder with coordinate-bias embedding, enabling scale-adaptive representation learning. Furthermore, we establish UAV-TSR—the first cross-scene benchmark dataset specifically designed for UAV-based thermal SR. Extensive experiments demonstrate that our method consistently outperforms state-of-the-art approaches across the full spectrum of scaling factors, significantly improving detail fidelity and edge sharpness in reconstructed thermal images. Both source code and the UAV-TSR dataset are publicly released.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multimodal Learning

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Thermal imaging can greatly enhance the application of intelligent unmanned aerial vehicles (UAV) in challenging environments. However, the inherent low resolution of thermal sensors leads to insufficient details and blurred boundaries. Super-resolution (SR) offers a promising solution to address this issue, while most existing SR methods are designed for fixed-scale SR. They are computationally expensive and inflexible in practical applications. To address above issues, this work proposes a novel any-scale thermal SR method (AnyTSR) for UAV within a single model. Specifically, a new image encoder is proposed to explicitly assign specific feature code to enable more accurate and flexible representation. Additionally, by effectively embedding coordinate offset information into the local feature ensemble, an innovative any-scale upsampler is proposed to better understand spatial relationships and reduce artifacts. Moreover, a novel dataset (UAV-TSR), covering both land and water scenes, is constructed for thermal SR tasks. Experimental results demonstrate that the proposed method consistently outperforms state-of-the-art methods across all scaling factors as well as generates more accurate and detailed high-resolution images. The code is located at https://github.com/vision4robotics/AnyTSR.

Problem

• Enhance low-resolution thermal UAV images flexibly
• Overcome fixed-scale super-resolution limitations efficiently
• Reduce artifacts in thermal image upsampling

Innovation

• Any-scale thermal SR within single model
• Image encoder with specific feature code
• Coordinate offset embedded upsampler