DeepLight: Reconstructing High-Resolution Observations of Nighttime Light With Multi-Modal Remote Sensing Data

TL;DR

This paper introduces DeepLight, a novel method and dataset for reconstructing high-resolution nighttime light images from multi-modal remote sensing data, improving the utility of NTL observations for SDG assessment.

Contribution

The paper presents DeepLightMD, a multi-sensor dataset, and DeepLightSR, a calibration-aware super-resolution method that effectively fuses heterogeneous remote sensing data for high-resolution NTL reconstruction.

Findings

DeepLightSR outperforms 8 competing methods in PSNR and PIQE metrics.

The dataset and method enable more accurate and detailed nighttime light observations.

High-resolution NTL data can better support SDG progress assessment.

Abstract

Nighttime light (NTL) remote sensing observation serves as a unique proxy for quantitatively assessing progress toward meeting a series of Sustainable Development Goals (SDGs), such as poverty estimation, urban sustainable development, and carbon emission. However, existing NTL observations often suffer from pervasive degradation and inconsistency, limiting their utility for computing the indicators defined by the SDGs. In this study, we propose a novel approach to reconstruct high-resolution NTL images using multi-modal remote sensing data. To support this research endeavor, we introduce DeepLightMD, a comprehensive dataset comprising data from five heterogeneous sensors, offering fine spatial resolution and rich spectral information at a national scale. Additionally, we present DeepLightSR, a calibration-aware method for building bridges between spatially heterogeneous modality data in the multi-modality super-resolution. DeepLightSR integrates calibration-aware alignment, an auxiliary-to-main multi-modality fusion, and an auxiliary-embedded refinement to effectively address spatial heterogeneity, fuse diversely representative features, and enhance performance in $8\times$ super-resolution (SR) tasks. Extensive experiments demonstrate the superiority of DeepLightSR over 8 competing methods, as evidenced by improvements in PSNR (2.01 dB $ \sim $ 13.25 dB) and PIQE (0.49 $ \sim $ 9.32). Our findings underscore the practical significance of our proposed dataset and model in reconstructing high-resolution NTL data, supporting efficiently and quantitatively assessing the SDG progress.