摘要
为了研究复杂流场环境下的包含遮挡物的非完全数据层析重建问题,从莫尔层析的基本理论出发,提出了一种将包含先验知识的属性矩阵融入迭代过程的全新的基于级数展开类的莫尔层析迭代算法。在此基础上,通过数值模拟,重建了包含遮挡物的三峰高斯分布的温度场,取得了理想的重建结果,并在相同条件下与层析变换类算法中滤波反投影算法进行了对比。结果表明,将先验知识以属性矩阵的形式融入迭代过程后,新算法与滤波反投影算法相比,能有效地处理包含遮挡物的非完全数据重建问题,为莫尔层析应用于实际测量奠定基础。
For the propose of incomplete data reconstruction caused by opaque objects in a complex flow field, a novel iterative algorithm based on series expansion methods by Moire tomography is developed from the essence of Moire deflection, the property matrix of a priori knowledge is introduced into the iterative process. The algorithm is studied by using it to reconstruct three-hump simulated temperature fields with opaque objects, so an ideal reconstruction results are achieved. Results are also compared with results from the traditional filtered back-projection algorithm under same conditions. Results show that having been combined with the property matrix of a priori information, new algorithm is more efficient to handle incomplete data reconstructive problems than the filtered back-projection algorithm. So the foundation of taking Moire tomography into practice is established.
出处
《激光技术》
CAS
CSCD
北大核心
2007年第2期153-155,159,共4页
Laser Technology
关键词
信息光学
莫尔层析
重建算法
非完全数据
information optics
Moire tomography
reconstruction algorithm
incomplete data
