Long-range Turbulence Mitigation:
A Large-scale Dataset and A Coarse-to-fine Framework
🔥 Accepted by 2024ECCV

The proposed method CDSP could effectively handle long-range atmospheric turbulence with severe geometric distortions.

Abstract

Long-range imaging inevitably suffers from atmospheric turbulence with severe geometric distortions due to random refraction of light. The further the distance, the more severe the disturbance. Despite existing research has achieved great progress in tackling short-range turbulence, there is less attention paid to long-range turbulence with significant distortions. To address this dilemma and advance the field, we construct a large-scale real long-range atmospheric turbulence dataset (RLR-AT), including 1500 turbulence sequences spanning distances from 1 Km to 13 Km. The advantages of RLR-AT compared to existing ones: turbulence with longer-distances and higher-diversity, scenes with greater-variety and larger-scale. Moreover, most existing work adopts either registration-based or decomposition-based methods to address distortions through one-step mitigation. However, they fail to effectively handle long-range turbulence due to its significant pixel displacements. In this work, we propose a coarse-to-fine framework to handle severe distortions, which cooperates dynamic turbulence and static background priors (CDSP). On the one hand, we discover the pixel motion statistical prior of turbulence, and propose a frequency-aware reference frame for better large-scale distortion registration, greatly reducing the burden of refinement. On the other hand, we take advantage of the static prior of background, and propose a subspace-based low-rank tensor refinement model to eliminate the misalignments inevitably left by registration while well preserving details. The dynamic and static priors complement to each other, facilitating us to progressively mitigate long-range turbulence with severe distortions. Extensive experiments demonstrate that the proposed method outperforms SOTA methods on different datasets.


Benchmark Examples

The proposed dataset RLR-AT contains atmospheric turbulence with diverse distances and scenes.


Benchmark Collection and Statistics

Illustration of the proposed dataset RLR-AT. (a) Long-range imaging with larger focal length lens through turbulence. (b) Typical turbulence with diverse distances and scenes. (c) Statistics of distance and scene of the proposed benchmark.


Frequency-aware Reference Frame Vs Temperol Average Reference Frame

Comparison between Temp Avg and proposed Frequency-aware reference frame (FRF)


Visual Comparisons on RLR-AT

Comparisons of long-range turbulence mitigation at various distances on RLR-AT




Demo

Section 1: Examples of Real Long-range Atmospheric Turbulence Benchmark. Section 2: Performance of CDSP on real-world Turbulence RLR-AT. Please change the video quality to 1080p for better visualization.


Dataset Download

•RLR-AT (Small-version) 45 Typical Turbulence Videos with 1920*1080 resolution.
Google Drive
• RLR-AT (Full-version) 1500 Turbulence Videos with 1920*1080 resolution.
Coming soon

Citation

Acknowledgements

This work was supported by the National Natural Science Foundation of China under Grant 62371203. The computation is completed in the HPC Platform of Huazhong University of Science and Technology.

The website template was borrowed from Michaël Gharbi and Ref-NeRF.