Improving Effective Temporal Resolution of Schlieren Imaging: A Decomposition-Based Frame Interpolation Method

Modern high-speed cameras can record images at very high frame rates, although typically at the expense of a smaller image size due to memory limitations. Using this capability, a dataset capturing a full domain of interest can be recorded, alongside a second series of images that capture a sub-region at higher frame rate. A data-processing technique is developed that combines the spatial and spectral information of two such image sets to generate image reconstructions with the full image domain and higher temporal resolution. The method uses proper orthogonal decomposition of both image sets to extract the constituent spatial and spectral components. A mode-pairing operation enables the full-domain spatial modes to be matched to the corresponding high-frequency spectral content from equivalent flow structures. The method is applied to schlieren imaging data for three compressible flowfields: a Mach 1.3 underexpanded jet flow; incipient transonic buffet over an aerofoil at Mach 0.75; and the perturbed bow shock of a blunt body with a contra-flow nose-jet at Mach 2.0. When the effective frame rates of these image sets are doubled, approximately 20%, 45% and 60% of the pixel values in the temporally varying images are accurately reconstructed for the jet, buffet and bow shock images respectively.