三维层析诊断跨界面成像修正方法研究

Three-dimensional (3D) tomography simultaneously records two-dimensional projections of the target flame from different perspectives using high speed cameras, and introduces the projections into a tomography algorithm to reconstruct the 3D structure of the flame. This new technique is able to resolve the flame development with both temporal resolution and spatial resolution in all three aspects. However, in the flame measurements inside the engine cylinders, significant image distortion of the projections is recognized due to the light refraction through the cylinder wall. As a result, 3D tomography would have poor accuracy if the image distortion is not corrected. This work develops a cross-interface image distortion correction algorithm based on the parallel-beam hypothesis and Ray-tracing method using MATLAB. The algorithm is then validated using numerical simulation.