层析成像是恢复日冕三维密度结构的强大技术。在本文中，我们通过在正则化项中引入径向权重，提出了一种改进的层析成像方法。径向权重在不同高度上提供了平衡的密度值平滑处理，有助于恢复较低高度处更精细的结构，同时稳定解并防止较高海拔处出现振荡伪影。我们将该技术应用于重建卡林顿周期（CR）2098的三维电子密度，使用双星任务中航天器B上的内日冕仪（COR1）在两周内的偏振亮度（pB）观测数据，其中径向权重函数采用背景逆强度，通过平均所使用的全部pB图像计算得到。层析成像与磁流体力学（MHD）模拟之间在不同高度处的密度分布对比显示了良好的一致性。我们发现，径向加权不仅有效纠正了使用二阶平滑重建时靠近内边界处的过平滑效应，而且显著提高了使用零阶平滑重建的质量。此外，比较单卫星数据与多视角数据对CR 2091的重建结果表明，日冕演化和动力学可能对重建的密度结构有显著影响。这一改进的层析成像方法已被用于基于STEREO/COR1-B数据创建CR 2052至CR 2154期间（2007年1月8日至2014年9月17日）的三维密度数据库。

Tomography is a powerful technique for recovering the three-dimensional (3D) density structure of the global solar corona. In this work, we present an improved tomography method by introducing radial weighting in the regularization term. Radial weighting provides balanced smoothing of density values across different heights, helping to recover finer structures at lower heights while also stabilizing the solution and preventing oscillatory artifacts at higher altitudes. We apply this technique to reconstruct the 3D electron density of Carrington Rotation (CR) 2098 using two weeks of polarized brightness (pB) observations from the inner coronagraph (COR1) onboard spacecraft-B of the twin Solar Terrestrial Relations Observatory (STEREO), where the radial weighting function is taken as the inverse intensity background, calculated by averaging all the pB images used. Comparisons between density distributions at various heights from the tomography and magnetohydrodynamics (MHD) simulations show good agreement. We find that radial weighting not only effectively corrects the over-smoothing effect near the inner boundary in reconstructions using second-order smoothing but also significantly improves reconstruction quality when using zeroth-order smoothing. Additionally, comparing reconstructions for CR 2091 from single-satellite data with that from multi-viewpoint data suggests that coronal evolution and dynamics may have a significant impact on the reconstructed density structures. This improved tomography method has been used to create a database of 3D densities for CRs 2052 to 2154, based on STEREO/COR1-B data, covering the period from 08 January 2007 to 17 September 2014.