A Stochastic Visualization Technique for Noisy 3D Unsteady Flow-Direction Data
Noriyasu Omata, Susumu Shirayama · Kashika Jouhou Gakkai rombunshuu gappon/Kashika Jouhou Gakkai rombun gapponshuu/Kashika Jouhou Gakkai rombunshuu · 2017
While quantitative flow data are subjected to qualitative analysis by their "appearance", it has recently been suggested that quantitative analysis can be performed also from qualitative experiments. In this paper, we propose a method to quantitatively analyze and visualize spatiotemporal data acquired from the tuft method, which is regarded as lacking quantitativity and objectivity. We propose a method to obtain the temporal division and spatial division through time by machine learning methods using stochastic model. By applying this method to the actual data, we succeeded in extracting the temporal and spatial pattern and showed that the proposed method has certain validity.