Abstract
A new nonlinear tomographic flow visualization technique for use in limited data situations is developed using techniques from the emerging field of evolutionary computing. The new technique uses both pure and hybrid genetic algorithms from the field of evolutionary computing. For verification of the technique, both axisymmetric and asymmetric phantom density fields are tested for a limited number of projections under interferometric visualization. For the hybrid genetic algorithm, investigations are conducted to determine the appropriate degree of hybridization for use in flow visualization. This new method can provide an alternative to common reconstruction algorithms for use in limited data cases as well as provide a direct reconstruction method for data generated from nonlinear projection mechanisms.
Lyons, Donald Paul (1997). Genetic algorithm based tomographic flow visualization. Master's thesis, Texas A&M University. Available electronically from
https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1997 -THESIS -L96.