Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge
summary
The gist
Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge demonstrates that employing the Lens Aperture as a Knife-Edge (LAKE) with two identical LEDs has the…
In short
The research introduces a method for bidirectional schlieren imaging using two identical LEDs and a lens aperture as a knife-edge. This setup allows controlling LED brightness to create weighted sums of horizontal and vertical cutoff images on each frame. Postprocessing these alternating monochrome images results in a colorized bidirectional schlieren image, enabling the visualization of flow gradients in both directions.
Key concepts
- Schlieren Imaging
- Schlieren imaging visualizes fluid flow by detecting variations in the refractive index or density of the medium. Traditionally, a knife edge is used to translate these density gradients into visible intensity changes on an image plane. This technique is essential for visualizing flow structures.
- Dual-LED Illumination Setup
- The system uses two surface-mount LEDs arranged in an L-shape to provide illumination for both horizontal and vertical cutoff visualization. One LED controls the horizontal cutoff, and the other controls the vertical one, allowing independent control over these two directional visualizations.
- Temporal Multiplexing
- Since a single monochrome camera cannot capture both cutoffs simultaneously, this technique alternates between acquiring horizontal and vertical cutoff images for each camera frame. This sequence of alternating captures is then used in postprocessing to create the final, composite colorized bidirectional schlieren image.
Terminology used across episodes
This episode discusses
- Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge · Paper Radio
The paper
Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge · Read on arXiv
VIMOD KUMAR, MANISH KUMAR
Centre for Sensors, Instrumentation and Cyber-Physical System Engineering Indian Institute of Technology Delhi
DOI: 10.1364/OL.616609
Transcript
Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.
Kai: I'm Kai.
Mira: And I'm Mira, here with Lev.
Kai: We're looking at the paper: "Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge".
Mira: Here's the gist of the paper. Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge demonstrates that employing the Lens Aperture as a Knife-Edge (LAKE) with two identical LEDs has the distinct advantage of obtaining schlieren images along two cutoff orientations.
Kai: This method allows for controlling LED brightness to obtain a desirable weighted sum of both horizontal and vertical schlieren cutoff images on each camera frames.
Mira: Fundamental Concept of Bidirectional Cutoff Schlieren imaging is a well-known method to visualize flow in transparent media.
Kai: Traditional schlieren employs a knife edge to translate a density gradient in the flow into an intensity variation on the image plane.
Paper summary: Kai: So we're looking at this paper, "Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge". What they're proposing is how you can get those two different cutoffs—horizontal and vertical—at the same time.
Mira: Right, it sounds like the authors are showing that using two identical LEDs with this specific setup gives them an advantage for getting those images along both cutoff orientations. The core claim is about controlling the LED brightness to get a weighted sum of both horizontal and vertical schlieren cutoff images on each camera frame.
Lev: And what they're claiming is that this dual-LED approach, using what they call the Lens Aperture as a Knife-Edge or LAKE, lets you switch between horizontal and vertical cutoffs without any crosstalk in the sequence. The setup involves one LED for horizontal and another for vertical cutoff, separated by a distance of two point eight millimeters.
Kai: It's interesting how they're controlling that brightness to get this weighted sum. But what does that actually mean for getting a real bidirectional schlieren image?
Mira: Well, the paper is clear that what they produce isn't a true bidirectional schlieren image right away. It’s a sum of horizontal and vertical cutoff images, but they show how you can use postprocessing to turn that monochrome sequence into a colorful one.
Lev: That postprocessing part is important because it shows you how to actually get the color visualization from the data they are collecting. They use things like 'lighten' and 'exclusion' blending modes in software like Adobe Photoshop or DaVinci Resolve, which operate pixel by pixel.
Kai: So, they build this hardware setup with an Arduino board to control everything. How fast can this thing actually take pictures? That seems important for a real experiment.
Lev: They achieved an imaging speed of five hundred frames per second with the camera hardware they were using at the time, which is pretty fast for these kinds of setups. They use digital synchronization between the camera frames and the LEDs to make sure you don't get crosstalk when switching orientations, which is a key technical detail.
Mira: The paper mentions that this method doesn't rely on a single light source being reshaped with slits or masks or split by a beam splitter, which is what all existing bidirectional schlieren techniques do. That’s a bit of an interesting design choice there.
Kai: It seems like they are trying to simplify the setup by not needing that extra light manipulation, which would be complicated. But what about the math behind how the cutoffs combine?
Mira: They point out that changing cutoff orientation follows a vector-like sum, where the length of that vector represents the strength of the LED and its direction represents the cutoff’s orientation. But they also note that this similarity to a vector sum stops at just finding the resultant vector orientation and doesn't account for resultant vector length considerations.
Paper summary: Lev: That means you get an image with less contrast because there's this signal averaging effect when you add those cutoffs together, which is something you have to account for in the final interpretation of the data.
Kai: So, what are the practical results they show from their experiments? What does this mean for someone just looking at a flow visualization sequence?
Mira: The middle image in their sequence, where they use equal contributions from both horizontal and vertical cutoffs, appears to provide a bidirectional schlieren image of the flow. And when you zoom in on those images, they say the sum image preserves both horizontal and vertical flow boundaries.
Lev: That preservation of boundaries is what makes it useful for visualizing the structure of the flow itself, even though it's not perfectly pure bidirectional schlieren yet.
Kai: It’s clear they are focused on getting a result that looks like a bidirectional image, even if it needs that postprocessing step to look right. So, what's the big picture here for flow visualization?
Mira: The implication is that you don't need super complicated optical setups with multiple light sources to get both horizontal and vertical gradient information simultaneously. This approach lets you control the relative strengths of those two signals directly through the LED brightness adjustment, which simplifies the hardware design considerably.
Lev: From an engineering standpoint, it's feasible because surface-mount LEDs are bright enough to record well-exposed images even with a six hundred microsecond exposure time. The distance between the source LEDs is two point eight millimeters, and that’s smaller than half the radius of curvature of their mirror setup.
Kai: So they're suggesting this is practical for getting quantitative measurements of density in flow, not just pretty pictures. What are they saying about extending this work?
Mira: They are confident that this approach can be extended to a quantitative measurement of density in flow, which is a big step beyond just visualizing the boundaries. The paper also mentions that they have filed a patent application for this specific approach.
Lev: And one limitation they state is that achieving a true bidirectional schlieren image with a monochrome camera only happens when you isolate those horizontal and vertical cutoff images from one another, which requires their temporal multiplexing approach to alternate acquisition.
Kai: So to wrap up, this paper is about using dual LEDs and a lens aperture as a knife-edge to get both horizontal and vertical schlieren cutoffs at once by controlling brightness, resulting in a weighted sum image that can be colorized later.
Mira: It's an interesting technique because it shows you how you can use simple illumination control to manage the sensitivity of your flow visualization system across two dimensions.
Lev: It’s feasible with existing hardware and it gives a clearer path toward quantitative density measurements, provided you handle that signal averaging effect in your analysis.
Kai: That's what we have here from "Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge".
Conclusion: Kai: So we've seen how they built this setup using two LEDs to get both horizontal and vertical cutoffs on one frame at a time, right?
Mira: Yeah, the title itself is pretty descriptive, "Bidirectional schlieren imaging using dual-LED illumination and a lens aperture as a knife-edge." It just tells you exactly what the apparatus is doing.
Lev: From an error correction standpoint, I'm interested in how they managed that switching between horizontal and vertical states without introducing crosstalk on the camera frames. That synchronization mechanism sounds tricky to build reliably on actual hardware.
Kai: It is tricky because they are relying on digital timing between the Arduino board and the LEDs, using a two thousand microsecond duration for each state change. I wonder what kind of noise that introduces into the final image data.
Mira: Well, the implication here is moving away from needing entirely separate light sources or complex beam splitters just to see both flow directions at once. It simplifies the optical path significantly.
Lev: But you have to remember, they aren't getting a true bidirectional schlieren image right away; they’re getting a weighted sum of two separate ones that need color postprocessing. That averaging effect is something we have to watch out for when we think about quantitative density measurements later on.
Kai: Right, so the authors are confident this approach can actually be extended to measuring flow density quantitatively, not just visualizing boundaries. That's a big step beyond just seeing where the flow is going.
Mira: It suggests that if you can control those LED brightness levels precisely, you gain a lot of flexibility in how you sample the gradient information across the two dimensions.
Lev: The hardware seems feasible because surface mount LEDs are bright enough for six hundred microseconds exposure times, which is good for keeping the measurement record exposed without needing super high-power laser sources.
Kai: So, they're showing a practical way to get that simultaneous view using existing components and simple electronics. It’s about making complex visualization accessible.
Mira: And they've filed a patent for this specific approach, which shows the authors are serious about the underlying hardware design being novel. That’s important context for how we see this technology developing in the field.
Lev: So, while it’s a solid visualization tool, we need to keep an eye on that signal averaging effect they mentioned when moving into real quantitative measurements of flow density.
Kai: Exactly. We need to know if that averaged image actually gives us the correct physical parameters about the fluid flow.
More episodes
- 2611.14082-Two atoms with strong light-induced dipolar interactions
- 2611.14040-Characterization of BNLs 20 and 30 micrometer thick LGADs for the Electron Ion Collider
- 2611.13399-Lift-element-based dynamic network of vortex gust-airfoil interactions
- 2611.13476-Hopfion Superlattices in 3D Twisted Moir'e Light Fields
- 2611.13642-Investigation of the timing performance of double-sided p-type 3D silicon sensors produced by CNM
- 2611.13830-SiPM optical readout of a PICOSEC Micromegas precise timing detector
- 2611.14152-Where are the shortest-path nodes in spatial networks?
- 2611.14219-Sub-Ice: semi-automated mapping of basal channels from ice shelf surface elevation data
- 2611.16341-SUMNO: Spatially Unrolled Multigrid Neural Operator for Fluid Flow Surrogates
- 2611.16090-Multi-UTE Myelin Fraction Mapping: Repeatability and Cross-Field-Strength Reproducibility at 1.5T and 3T