Markov Chain and Monte Carlo Predictions for Light Multiple Scattering Applications

摘要

Multiple scattering is considered both an interference source and tentatively the information to measure the optical properties of mediums in many engineering applications, such as fuel spray atomization in IC engines, medical diagnostics for bio-tissues, and atmosphere detection, etc. However, the understanding of multiple scattering is hindered by the stochastic nature of random scattering, especially in the intermediate regime. This work reports a Markov Chain solution to analyze the angular distribution of transmitted photons and compared against a typical method, Monte Carlo algorithm. The Markov Chain method is then utilized to perform an inversion process to derive the optical properties inside the medium and various reconstruction algorithms were tested. Results have shown that Markov Chain method is a viable technique in reducing the computational cost yet preserving the computation fidelity.