Abstract
We propose a minimum variation of solution method to determine the optimal regularization parameter for singular value decomposition for obtaining the initial distribution for a Chahine iterative algorithm used to determine the particle size distribution from photon correlation spectroscopy data. We impose a nonnegativity constraint to make the initial distribution more realistic. The minimum variation of solution is a single constraint method and we show that a better regularization parameter may be obtained by increasing the discrimination between adjacent values. We developed the S-R curve method as a means of determining the modest iterative solution from the Chahine algorithm. The S-R curve method requires a smoothing operator. We have used simulated data to verify our new method and applied it to real data. Both simulated and experimental data show that the method works well and that the first derivative smoothing operator in the S-R curve gives the best results.
© 2012 Optical Society of America
Full Article | PDF ArticleMore Like This
Xinjun Zhu, Jin Shen, Wei Liu, Xianming Sun, and Yajing Wang
Appl. Opt. 49(34) 6591-6596 (2010)
Janusz Mroczka and Damian Szczuczyński
Appl. Opt. 51(11) 1715-1723 (2012)
Yajing Wang, Jin Shen, Liu Wei, Zhenhai Dou, and Shanshan Gao
Appl. Opt. 52(12) 2792-2799 (2013)