Nighttime road lighting is crucial for transportation and substantially contributes to power consumption. To enhance energy efficiency, numerous Light Emitting Diode (LED) lamps have been deployed across urban road networks. Assessing their effectiveness, however, has been challenging due to the coarse spatial resolution of traditional glimmer imagery. This study leverages high-resolution imagery from the newly launched SDGSAT-1 satellite, equipped with a Glimmer sensor, to quantitatively evaluate the impact of LED integration on power conservation within urban road networks. The SDGSAT-1 satellite provides unprecedented clarity with 10 m panchromatic and 40 m multi-spectral RGB resolutions, enabling a detailed analysis of illuminated road networks and the differentiation of LED and non-LED lighting sources. We utilized an unsupervised machine learning approach to extract and categorize lighting networks from panchromatic images based on spectral characteristics in RGB images, achieving an F1-measure of up to 98.11% after field validation. Our results reveal substantial urban nightlife vibrancy and effective energy-saving strategies in the Guangdong-Hong Kong-Macao Greater Bay Area and the Yangtze River Delta, in contrast to older, economically developed cities where conservation efforts were less effective. This study underscores the potential of SDGSAT-1 imagery for precise nighttime lighting assessments and offers valuable insights for optimizing urban development and energy conservation policies.

中文翻译：

---

基于SDGSAT-1微光影像的路网节电评估


夜间道路照明对于交通至关重要，并且会极大地增加电力消耗。为了提高能源效率，城市道路网络中部署了大量的发光二极管 (LED) 灯。然而，由于传统微光图像的空间分辨率较差，评估其有效性一直具有挑战性。本研究利用新发射的配备 Glimmer 传感器的 SDGSAT-1 卫星的高分辨率图像，定量评估 LED 集成对城市道路网络节能的影响。 SDGSAT-1 卫星提供前所未有的清晰度，具有 10 m 全色和 40 m 多光谱 RGB 分辨率，能够对照明道路网络以及 LED 和非 LED 照明源进行详细分析。我们利用无监督的机器学习方法，根据 RGB 图像的光谱特征从全色图像中提取照明网络并进行分类，在现场验证后实现了高达 98.11% 的 F1 测量。我们的研究结果显示，粤港澳大湾区和长三角地区的城市夜生活充满活力，节能策略也有效，而相比之下，经济发达的老城市的节能工作效果较差。这项研究强调了 SDGSAT-1 图像在精确夜间照明评估方面的潜力，并为优化城市发展和节能政策提供了宝贵的见解。