摘要
研究非完全数据光偏折层析重建技术,提出了基于偏折角修正迭代的重建算法族,并结合抑制噪声、减弱重建模糊性和边缘效应、优化松弛系数等发展出多种重建迭代形式,使重建精度得到有效提高。在有限角投影、随机噪声数据和无平滑处理等重建条件下,应用该族算法对复杂温度分布进行数值重建,对不同算法的重建效能和重建误差进行对比,分析了各算法与相关的改进代数重建技术在重建性能上的明显差异,并得出最优的偏折层析重建算法。
A series of deflection angle revision reconstruction algorithms are developed for limited-data deflection tomography.The algorithms are derived from the basic deflection formula and the deflection angles are used directly in iteration.In order to improve the reconstructed accuracy,the modified theories for algebraic reconstruction technique are applied to deflection tomographic reconstruction.A number of iterative methods are presented to suppress noise,solve edge distortion,and optimize relaxation factor.The reconstruction techniques are tested using simulated data for incomplete projection situation with random noise.The reconstructed results by different algorithms are analyzed,and the reconstructed errors are discussed.The best-performing algorithm for deflection tomography is obained.
出处
《光电子．激光》
EI
CAS
CSCD
北大核心
2010年第6期911-916,共6页
Journal of Optoelectronics·Laser
基金
山东省中青年科学家科研奖励基金资助项目(2008BS01004)
关键词
信息光学
偏折层析
重建算法
非完全数据
information optics
deflection tomography
reconstruction algorithm
incomplete projection data
