Application of Conditioned Level-Set Method to OH-PLIF Image Processing of Typical Molecule Diffusion Flames

Abstract

A novel approach for OH-PLIF image processing, namely conditioned level-set (CLS) method, is presented in this paper. The complete CLS method consists of raw image pre-processing, adaptive intensity shareholding for linearization, initial contour extraction, region-lock optimization and level set iteration. The successful application of this method to typical molecule diffusion flames demonstrates its robustness to noise sensitivity and flame topology, and its outperformance over traditional edge detectors in terms of flame contour identification. The CLS method provides possibility to further image analysis such as mean reaction zone tracking. Based on the results from RANS simulation with adequate modelling, a novel numerical-aided approach for concentration calibration is also proposed. The CLS method can readily be employed to experimental images obtained from other planar scalar measurements. The proposed method has potential to be integrated to image processing modules of scientific software.