跳转至

利用空间望远镜数据与生成式AI增强数字巡天的成像能力

文章背景与核心概要

数字巡天项目能够提供极高的图像数据吞吐量并覆盖广阔的天区,但在分辨率和成像细节上通常落后于空间望远镜。相比之下,空间望远镜擅长深空成像,却缺乏地面巡天系统的高效吞吐能力。为了弥合这一差距,本文引入了一种创新的生成式AI驱动方法,通过在空间望远镜图像上训练生成模型并利用星系形态固有的结构模式,有效地将地面巡天获取的微弱信号转化为清晰、细节丰富的星系图像。

研究团队开源了整个处理流程、成对的训练数据集、包含63,202张增强星系图像的星表,以及封装了定制生成式AI模型的软件工具。这项工作成功将地面巡天的高吞吐量与空间望远镜的高图像质量结合起来,为现代天体物理学研究提供了强大的数据增强手段。


摘要 (Summary)

Digital sky surveys provide high data throughput and cover vast footprints, but their resolution and imaging power generally lag behind space-based telescopes. Conversely, space telescopes excel at deep-universe imaging but lack the rapid survey throughput of ground-based systems.

This paper introduces a novel generative AI-driven methodology to bridge this gap. By training generative models on space telescope imagery and leveraging the inherent structural patterns of galaxy shapes, the approach effectively transforms weak signals from ground-based surveys into sharp, detailed galaxy images. The authors provide an open-source pipeline, paired training datasets, a catalog containing 63,202 enhanced galaxy images, and a software tool encapsulating the custom generative AI model.

数字巡天虽然能够提供极佳的图像数据吞吐量并覆盖大片天区,但其成像能力通常逊色于空间望远镜。另一方面,空间望远镜具备出色的成像能力并能够拍摄深空,却无法提供与先进地面巡天相当的吞吐量。

本文引入了一种新颖的生成式AI驱动方法来弥合这一差距。通过在空间望远镜图像上训练生成模型,并利用星系形态固有的结构模式,该方法能够有效地将地面巡天的微弱信号转化为清晰、细节丰富的星系图像。该方法将地面巡天的高吞吐量与空间望远镜的图像质量完美结合。作者开源了该方法的源代码、成对的训练数据、一个包含63,202张经该方法增强的星系图像星表,并提供了一个封装了完整流程与定制生成式AI模型的软件工具,以便生成画质增强的星系图像。


文章详情 (Article Details)

  • arXiv Identifier: arXiv:2608.20666 [astro-ph.IM]
  • Authors: Sai Teja Erukude, Lior Shamir
  • Submitted On: 21 August 2026
  • Status: Accepted in MNRAS
  • Primary Subject: Instrumentation and Methods for Astrophysics (astro-ph.IM)
  • Secondary Subjects: Astrophysics of Galaxies (astro-ph.GA), Artificial Intelligence (cs.AI), Machine Learning (cs.LG)
  • License: Creative Commons Attribution 4.0 license icon
  • arXiv 标识符: arXiv:2608.20666 [astro-ph.IM]
  • 作者: Sai Teja Erukude, Lior Shamir
  • 提交时间: 2026年8月21日
  • 状态: 已被 MNRAS 接收
  • 主学科: 天体物理学仪器与方法 (astro-ph.IM)
  • 次学科: 星系天体物理学 (astro-ph.GA)、人工智能 (cs.AI)、机器学习 (cs.LG)
  • 许可协议: 知识共享署名 4.0 license icon

摘要正文 (Abstract)

While digital sky surveys provide excellent throughput of image data and can cover a large footprint, their imaging power is normally inferior to that of space-based telescopes. Space-based telescopes, on the other hand, provide excellent imaging power and can image the deep Universe, but cannot provide the same throughput as advanced ground-based sky surveys.

Here, we utilize generative AI to elevate the quality of galaxy images taken by ground-based telescopes to the level of details enabled by space telescopes. The solution is based on the nature of galaxy shapes, allowing generative AI trained on space-based images to convert weak signal into detailed and clear galaxy images. The method allows for combining the high throughput of ground-based sky surveys with the image quality of space-based telescopes. The source code for the method is available, as well as paired training data and a catalog of 63,202 galaxy images enhanced by the proposed method. We also provide a software tool that encapsulates the entire pipeline and the custom generative AI model to generate galaxy images with enhanced quality.

虽然数字巡天提供了极佳的图像数据吞吐量并能覆盖广阔的天区,但其成像能力通常不如空间望远镜。而空间望远镜则拥有卓越的成像能力并能拍摄深空,但无法提供与先进地面巡天相同的高吞吐量。

在这里,我们利用生成式AI将地面望远镜拍摄的星系图像质量提升至空间望远镜所能达到的细节水平。该解决方案基于星系形态的本质,使在空间图像上训练的生成式AI能够将微弱信号转换为详细且清晰的星系图像。该方法将地面巡天的高吞吐量与空间望远镜的图像质量结合在了一起。本研究提供了该方法的源代码、成对的训练数据以及一个包含63,202个通过该方法增强的星系图像的星表。我们还提供了一个软件工具,其中封装了完整的处理流程和定制的生成式AI模型,用于生成画质增强的星系图像。


获取与资源 (Access and Resources)