This paper gives an efficient approach to reconstruct moving multiple objects （multi-object）. Each object has independently rigid motion which includes translation and rotation. The traditional FBP algorithm can resolve the one-object motion problem rather well. However, it suffers from perceptible motion artifacts in multi-object cases. This paper proposes a new motion-compensated reconstruction approach with a pdori knowledge of the rigid motion model. Both an FBP-type and an ART-type algorithm were derived. In an effort to evaluate the proposed algorithms, we have performed numerical studies by using different rigid motion models. Quantitative results demonstrate that the proposed FBP-type and ART-type algorithms can recon- struct multi-object free of motion artifacts.
A three-dimensional （3-D） phantom for the density distribution of the plasmasphere is established. The imaging processes of the extreme ultraviolet （EUV） Imager are computer-simulated, in which the Earth shelter is treated as a main problem. A modified ART method is devised to resolve the incomplete data reconstruction problem to validate and evaluate the proposed methods. The cone-beam EUV data are simu- lated based on the 3-D phantom from both a circular and semi-circular trajectories. Quantitative reconstruc- tion results demonstrate the correctness of the proposed modified ART algorithm. The CT technique can be used to calculate the global density of the plasmasphere from the EUV data.
在某些用于在线控制的CT系统中,经常有必要在扫描过程中检查中间结果从而排除故障。然而实时的重建与体绘制对系统计算能力的要求非常高,难以实现。本文针对锥束CT提出了一整套实时重建可视化方法,并在最新的图形处理器Tesla C 1060上实现。与传统的工作方式不同,这种方法在扫描得到每一张投影后,实时显示重建过程的中间结果。通过这种方法,可以很清楚地观察到投影数据逐步合成三维重建结果的过程,也可以预先看到探测器的缺陷对重建结果造成的破坏。同时,这种方法也是实现4D CT可视化的一种有效途径。
