利用CARTA中的体渲染技术探索天体物理数据立方体
Towards the exploration of astrophysical datacubes through volumetric rendering within CARTA
浏览论文内容
中文总结 AI 辅助
该研究在天文软件CARTA中开发了基于ThreeJS的可扩展交互式三维可视化方法，可本地可视化最大128GB的天体物理数据立方体，为大型数据立方体的三维可视化应用奠定基础。
中文摘要 AI 辅助
天体物理观测的数据产品可能包含三维或更多维度：最常见的是两个空间维度和第三个光谱维度，还可能存在偏振或时间等额外轴。三维图像数据的可视化对分析各类天体的运动学和演化尤为有用；然而，下一代观测站产生的数据规模庞大，且在二维屏幕上显示三维数据的复杂性使得可视化工作极具挑战性。本文中，我们在天文软件CARTA中提出了一种可扩展且交互式的三维可视化方法。我们使用ThreeJS及已知技术实现了一个原型三维渲染部件，并通过压缩、降采样等技术提升效率，发现最大达128GB的大型数据立方体可在本地可视化，加载时间仅需数分钟。我们分析了前端和后端的性能，发现实现的瓶颈在于前端，原因是前端将数据存储在单个数组中，而后端则以分块形式访问和传输数据。本研究通过在CARTA未来版本中实现三维渲染部件，为社区提供大型数据立方体的三维可视化铺平了道路。
英文摘要
Data products from astrophysical observations may contain three or more dimensions: most commonly two spatial dimensions and a third spectral dimension, and possibly additional axes such as polarisation or time. Visualising three-dimensional image data is particularly useful for analysing the kinematics and evolution of various objects; however, the large scale of data produced by next-generation observatories and the complexity of displaying 3D data on a 2D screen make visualisation challenging. In this article, we present a scalable and interactive 3D visualisation approach within the astronomy software CARTA. We implemented a prototype 3D rendering widget with ThreeJS and known techniques to improve efficiency (e.g. compression, downsampling) and found that large datacubes up to 128 GB can be visualised locally with a loading time of a few minutes. We analysed the performance in the frontend and in the backend, finding that the bottleneck of our implementation lies in the frontend. This is due to storing the data in a single array in the frontend, unlike in the backend, where it is accessed and sent in blocks. This study paves the way for making 3D visualisation of large datacubes available to the community by implementing a 3D rendering widget in a future release of CARTA.
