标题： 提升ASO-S任务中Lyman-α太阳望远镜（LST）的太阳圆盘成像仪（SDI）的图像质量 显示英文标题

标题： Improving image quality of the Solar Disk Imager (SDI) of the Lyman-alpha Solar Telescope (LST) onboard the ASO-S mission

摘要： 先进太阳观测卫星（ASO-S）任务中日冕仪（LST）的关键仪器太阳盘成像仪（SDI）的在轨校准和性能显示其空间分辨率远低于预期。 本文开发了SDI点扩散函数（PSF）和图像双变量优化算法（SPIBOA），以提高SDI图像的质量。 双变量优化方法巧妙地结合了深度学习与光学系统建模。 尽管缺乏SDI实际拍摄的真实图像及其光学系统函数的信息，该算法仍能有效从大量观测数据样本中直接估算出SDI成像系统的PSF。 我们使用估算的PSF对观测到的SDI图像进行反卷积校正，结果显示校正后的图像的空间分辨率比原始观测值提高了三倍以上。 同时，我们的方法还显著减少了观测到的SDI图像中的固有噪声。 SPIBOA现已成功集成到常规SDI数据处理中，为基于数据的科学研究提供了重要支持。 SPIBOA的发展和应用也为识别天文望远镜系统和提升观测图像质量开辟了新途径。 文中还讨论了在应用SPIBOA方法时的一些关键因素和注意事项。 显示英文摘要

摘要： The in-flight calibration and performance of the Solar Disk Imager (SDI), which is a pivotal instrument of the Lyman-alpha Solar Telescope (LST) onboard the Advanced Space-based Solar Observatory (ASO-S) mission, suggested a much lower spatial resolution than expected. In this paper, we developed the SDI point-spread function (PSF) and Image Bivariate Optimization Algorithm (SPIBOA) to improve the quality of SDI images. The bivariate optimization method smartly combines deep learning with optical system modeling. Despite the lack of information about the real image taken by SDI and the optical system function, this algorithm effectively estimates the PSF of the SDI imaging system directly from a large sample of observational data. We use the estimated PSF to conduct deconvolution correction to observed SDI images, and the resulting images show that the spatial resolution after correction has increased by a factor of more than three with respect to the observed ones. Meanwhile, our method also significantly reduces the inherent noise in the observed SDI images. The SPIBOA has now been successfully integrated into the routine SDI data processing, providing important support for the scientific studies based on the data. The development and application of SPIBOA also pave new ways to identify astronomical telescope systems and enhance observational image quality. Some essential factors and precautions in applying the SPIBOA method are also discussed.