Abstract
A series of previous studies, both numerical and experimental, have demonstrated the advantages of hyperspectral tomography (HT) as a promising technique to measure the two-dimensional distributions of temperature and species concentration in reacting flows. This paper intends to prepare the mathematical groundwork for extended use of the HT technique for three-dimensional and/or time-correlated measurements. Direct application of the methods developed previously encounters both experimental and computational difficulties. Numerical studies reported in this paper suggest that the use of proper orthogonal decomposition (POD) is effective to overcome these difficulties. The use of POD in HT significantly reduces the computational cost, enhances the fidelity of the tomographic reconstructions, and improves the stability of the reconstruction in the presence of measurement noise. Implications of these results for practical applications are also discussed.
© 2010 Optical Society of America
Full Article | PDF ArticleMore Like This
Lin Ma and Weiwei Cai
Appl. Opt. 47(21) 3751-3759 (2008)
Xinliang An, Thilo Kraetschmer, Kuya Takami, Scott T. Sanders, Lin Ma, Weiwei Cai, Xuesong Li, Sukesh Roy, and James R. Gord
Appl. Opt. 50(4) A29-A37 (2011)
Lin Ma and Weiwei Cai
Appl. Opt. 47(23) 4186-4192 (2008)