摘要
交错格采样是二维平行束扫描模式下的一种非传统采样形式,该采样形式下的投影数据带宽与传统采样的投影数据带宽相近.引入交错格采样的思想,对二维平行束扫描模式下传统采样的投影数据进行归组分块,块内数据具有交错格采样数据的特点,也具有交错格采样数据的频谱特征,保留了所有投影数据完整的频谱信息,一定程度上改善了迭代过程中的误差分配.由此对块迭代算法进行改进,形成交错格块迭代算法.实验结果表明,在相同迭代次数下,较之ART图像重建算法,交错格块迭代算法能有效抑制误差,提高成像精度.
Interlaced lattice sampling is a kind of non-traditional sampling form of two-dimensional parallel beam scanning mode. The projection data bandwidth of interlaced lattice sampling is close as the traditional sampling. The idea of interlaced lattice sampling was introduced to block the traditional sampiing projection data of two-dimensional parallel beam. The data of this block has the characteristics of interlaced lattice sampling data, and also has frequency spectrum characteristics of interlaced lattice sam- pling data. The data of this block reserves complete frequency spectrum information of all projection da- ta, and it improves the distribution of the error during the iteration to some extent. Thus an interlaced lattice block iterative algorithm has formed to improve block iterative algorithm. The experimental results show that, under the same number of iterations, compared with ART image reconstruction algorithm, interlaced lattice block iterative algorithm can effectively restrain the error and improve the imaging accuracy. Key words： image reconstruction; spectrum analysis of projection data; interlaced lattice sampling method; interlaced lattice block; block iterative
出处
《中北大学学报（自然科学版）》
CAS
北大核心
2016年第1期67-75,共9页
Journal of North University of China(Natural Science Edition)
基金
国家自然科学基金项目(61372150
61271425)
关键词
图像重建
投影数据频谱分析
交错格采样方法
交错格分块
块迭代
image reconstruction
spectrum analysis of projection data
interlaced lattice sampling method
interlaced lattice block
block iterative
