| Summary: | Schlieren photography is a technique commonly used for non-quantitative imaging of transparent fluids. The Schlieren system images differences in brightness by making use of a physical dependency of the refractive index on the density of fluids. Hence, Schlieren images show the density gradient in the probed measurement volume. [Settles, 2001] Flow phenomena imaged by optical Schlieren Systems such as conventional Z-type setups or focused Schlieren systems often involve three dimensional effects. [Ifti et al., 2018] Using common Schlieren setups either everything is integrated within the depth of field or only one specific 2D plane of the object is in focus and can be investigated. 3D information cannot be reconstructed in both cases. The goal of the current work is to reconstruct accurate 3D information from a single Schlieren image by using a plenoptic camera.
A plenoptic camera presents an advancement to a normal digital camera. The concept of this technique involves recording angular information about the incoming light rays in addition to the intensity distribution, hence a 4D Light field is captured. This is achieved by placing a microlens array in front of the image sensor to split every ray and therefore image every point in the object space on several pixels on the image sensor. As a result, spatial information in the axial depth of the measurement volume is captured by losing lateral resolution. [Ng et al., 2005] This 4D Light Field can then be used to obtain 3D information about the captured flow. For this purpose a reconstruction process has to be carried out as the real object has been convolved with the spatial impulse response of the camera system while the image was taken. The first option is to use the algorithm of a Limited-Angle-Tomography. Projecting the object space onto several pixels per point allows to reconstruct views from a finite number of angles. A set of these views can then be used as an input for a 3D image formation. A theoretically equivalent result can be obtained by a 3D Deconvolution. A key feature of a plenoptic camera is the ability to refocus an image to a number of focal depths after it has been taken. Each of these focal slices contains blurred features from other depths. By characterizing the impulse response of the plenoptic camera, the so called Point-Spread-Function this out-of-focus light can be removed. Aligning every focal slice delivers a 3D image of the flow object. [Levoy et al., 2006] [Bronxton et al., 2013]
In this paper a novel application of a plenoptic camera to a focused Schlieren system is presented. Compared to a normal camera this approach can also acquire axial information. Thus, a Schlieren system is enhanced by adding the third dimension to its imaging capability. In this project a Lytro Illum commercial plenoptic camera has been used to capture Light Field images with a focusing Schlieren benchtop system
|
|---|