Assessment of image processing methods to characterize the precise shape of the jet head in the hypersonic jet flows

摘要

Image-based methods are going to be the best ones in fluid dynamics, especially where other techniques are not capable to detect the phenomena of the flows. One of these topics is hypersonic jet flow, while recently several laboratory experiments of the long-term hypersonic jet flows have been performed to explore different morphological aspects of the flow, such as asymmetric stability properties, cocoon and pumping phenomena. In this paper, to get a deeper insight in the phenomenology of the hypersonic jets, a set of image processing methods has been applied to the laboratory fluorescent images of hypersonic jets, which helps to recognize the precise shape of the jet head. The captured images are pre-processed through nonlinear diffusion that removes the noises, while as the time elapses, the image is segmented into the jet and ambient media. Then, the edge detection operator, coupling with curve fitting function, is applied to specify the jet flow boundary.