Enhanced Background Oriented Schlieren (EBOS). (May 2019)

• Record Type: Journal Article
• Authors: Cozzi, F.; Göttlich, E
• Is Part Of: Journal of physics. Volume 1249(2019)
• Publication Date: 2019-05
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1249/1/012017
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: The Background Orientend Schlieren (BOS) is a line of sight optical technique that exploits light refraction arising in inhomogeneous density field to visualize or either measure density gradients. The working principle of BOS is quite simple: a background patter appears distorted when observed through an optical inhomogeneous media, by comparing the distorted and undistorted images the apparent local displacement of the background pattern can be measured. The images comparison is carried out using either cross-correlation or optical flow techniques. In the framework of geometric optics and for a 2D refractive index field the measured displacements is proportional to the refractive index gradient. When applicable, the Gladstone-Dale relationship provides the link between the refractive index and the fluid density. The uncertainty in the measured displacements affects the qualitative and quantitative results extracted from BOS images. Spatial and temporal filtering techniques can reduce the uncertainty at the expenses of the spatial and the temporal resolutions. This work describes a modified BOS technique, named Enhanced BOS (EBOS), which is able to reduce the displacement uncertainty without using any spatial or temporal filter. The EBOS technique uses N undistorted images of a grey-scale background pattern taken at slightly different potions, the N reference images are paired with a single distorted background image. Eventually displacement maps obtained from the N image couples are averaged together. The mean displacement map is an improved estimate of the actual displacement field. The EBOS technique can also be used with optical flow techniques anyhow we limited our investigation to image analysis based on cross-correlation technique. The paper reports the results obtained by applying EBOS and BOS techniques to the same test case. The comparison evidence the capability of EBOS to reduce displacement uncertainty respect to BOS by using a limited number of undistorted background images. EBOS and BOS techniques achieve almost the same uncertainty when the number of particles contained in an interrogation area is high, thus EBOS is most advantageous to use when small interrogation area are required.