Flame front detection and curvature calculation using level set

A level set method is proposed for flame front detection and curvature calculation on turbulent premixed OH Planar Laser Induced Fluorescence (OH-PLIF) images. The Flame front on an OH-PLIF image is detected by our proposed curve evolution model. In this model, a global term is designed to make the evolution curve converge to the flame front and a Gaussian kernel function is used to detect the flame front more accurately. For the detected flame front contour, we calculate its curvature using its level set function which represents the contour as the zero level of a higher dimensional embedding function. The precision of our method is verified by the comparison with a common curve-fitting method on synthetic images. The results on different flame images also demonstrate that our method can efficiently detect the flame fronts and get more accurate curvature on OH-PLIF images. The calculated curvature, as an important factor in understanding turbulence and chemistry interaction phenomena, can be used to perform quantitative analysis of flame fronts.