The split Bregman algorithm for temperature reconstruction based on the light-field imaging technique

The light-field imaging technique is a promising technique for the flame temperature measurement, and the reconstruction method plays a very important role in the application of this technique. This paper presents a numerical study on reconstruction of three-dimensional temperature distribution in participating media based on the light-field imaging theory. The Split Bregman algorithm is applied to the temperature reconstruction, and simulated exit radiation intensities on the boundary of participating media computing by the generalized sourced multi-flux method according to the simplified computational model of light-field imaging are employed as inputs for inverse analysis. Axisymmetric temperature distribution is reconstructed, and influences of measurement errors and radiative properties on the accuracy of estimation are also investigated. Results indicate that the temperature distribution can be reconstructed reasonably even with measurement errors, and the reconstruction accuracy decreases with the increase of extinction coefficient. Besides, a Non-axisymmetric temperature distribution is also reconstructed and results show that the temperature field can be estimated accurately when no measurement error exists. While noise exists, the reconstruction accuracy becomes lower. This study demonstrates that the Split Bregman algorithm is suitable for the temperature reconstruction based on the light-field imaging technique, and improvement should be made for the case of non-axisymmetric temperature distribution reconstruction.