Synthetic Light-in-Flight

TL;DR

This work introduces Synthetic Light-in-Flight (SLiF), a computational LiF framework that uses tunable continuous-wave lasers and standard CMOS cameras to synthesize fields at a synthetic wavelength $\Lambda$ from pairs of closely spaced wavelengths $\lambda_1$ and $\lambda_2$. By computationally aligning and summing multiple synthetic fields, SLiF creates a synthetic pulse train whose time-of-flight reveals depth and refractive-index information without pulsed lasers or high-speed detectors. Key contributions include the generation and phase alignment of synthetic fields, volumetric scene sectioning through scattering media, and post-acquisition temporal and spatial pulse shaping to visualize material properties and defects. The approach offers a flexible, speckle-robust alternative to traditional LiF and OCT/WLI, enabling high-resolution, full-field 3D imaging with potential applications in industrial inspection and biomedical imaging through challenging media.

Abstract

Light-in-flight (LiF) measurements enable the visualization of light paths through arbitrary, volumetric scenes, making light-matter interactions at ultrafast timescales visible. Traditionally, LiF measurements require specialized equipment, such as ultrashort pulse light sources and high-speed electronics, often limited by low spatial resolution. Herein, we introduce a novel computational approach,"Synthetic Light-in-Flight" (SLiF), that overcomes these constraints by relying solely on tunable, continuous wave (CW) lasers and off-the-shelf CMOS cameras. From multiple CW scene measurements at different optical wavelengths, we create multiple "synthetic fields," each at a "synthetic wavelength," which is the beat wave of two respective optical waves. These synthetic fields are robust to speckle and environmental fluctuations, enabling us to combine multiple synthetic fields into a "synthetic light pulse" that sections the volumetric scene. Additionally, we demonstrate that these complex synthetic pulse fields can be freely manipulated in the computer after their acquisition, allowing for spatial and temporal shaping of different sets of pulses from the same set of measurements to maximize the decoded information output for each scene. Finally, we show that the recovered time-of-flight information can be used to characterize physical scene properties, such as depth and refractive indices.

Figures (8)

• Figure 1: Synthetic Light-in-Flight (SLiF): Generation of synthetic pulses: a) Schematic of measurement setup: A scene is illuminated at multiple wavelengths and a holographic camera ballester.2024 captures the field information of each back-scattered optical field. b) Image of measured object. c-e) The set of captured optical fields (c) at different wavelengths $\lambda_{1}, ..., \lambda_{M}$ is used to create a set of synthetic fields (d) at different synthetic wavelengths $\Lambda_{1}, ..., \Lambda_{N}$ via computational pairwise mixing. The synthetic fields are computationally phase-aligned and superimposed to create a synthetic pulse train (e). The pulse becomes more well-defined, the more synthetic fields are added. f-h) Example images for: Speckled phase maps $\phi(\lambda_m)$ of captured optical fields (f). Phase maps $\phi(\Lambda_n)$ of calculated synthetic fields (g). Squared amplitude of assembled synthetic pulse $|P|^2$, shown at different time stamps between 0ps and 140ps. It can be seen that the assembled synthetic pulse precisely sections the object surface as it advances through the scene (see video https://drive.google.com/drive/folders/1rS9Itz3QuB3RZksTBMV8XeLqz7HhctGi?usp=drive_linkvideos.2024). The experimentally evaluated pulse FWHM for this measurement is $9.88\ \mathrm{ps}$ or $2.96\ \mathrm{mm}$.
• Figure 2: Synthetic Pulse Propagation Through a Volume (see videos https://drive.google.com/drive/folders/1rS9Itz3QuB3RZksTBMV8XeLqz7HhctGi?usp=drive_linkvideos.2024): a) Setup schematic: A collimated laser beam is emitted into a lightly scattering medium and reflected off a $45^\circ$ angled mirror. The holographic camera observes the scene normal to the light propagation plane. b) Image of the scene and embedded mirror from the camera's perspective and schematic light path. c) Computed synthetic pulse entering the scattering medium at $40\ \mathrm{ps}$. d) The pulse is computationally advanced and reflected off the mirror (dashed yellow) at around $153\ \mathrm{ps}$. e) The pulse continues propagation in reflected direction.
• Figure 3: Synthetic Pulse Propagation Through a Heavily Scattering Volumetric Medium (see videos https://drive.google.com/drive/folders/1rS9Itz3QuB3RZksTBMV8XeLqz7HhctGi?usp=drive_linkvideos.2024): a) Setup schematic: A collimated laser beam is emitted into strongly scattering medium with an embedded occluder. The camera observes the scene at a $90^\circ$ angle. b) Image of the scattering medium with embedded occluder. c) The computational pulse entering the scattering medium at $45\ \mathrm{ps}$. d) The pulse is computationally advanced and begins to spread, creating a photon horizon, blocked by the occluder (yellow dashed circle) at $113\ \mathrm{ps}$. e) Photon horizon traveling around the occluder back into a circular propagation path as it passes across the medium.
• Figure 4: Synthetic Pulse Propagation through a Scattering Phantom with Embedded Foreign Bodies: a) Schematic of object: Copper spheres with $4.5\ \mathrm{mm}$ diameter are embedded in a scattering phantom at various depths ($5\ \mathrm{mm}$–$20\ \mathrm{mm}$). b) Volumetric reconstruction of sphere locations from time-of-flight data of the SLiF measurement. c) A time slice highlighting the intensity of the reflection arriving from a single sphere at $28\ \mathrm{ps}$. d) Second time slice captured at $56\ \mathrm{ps}$ highlighting reflections from two metal spheres.
• Figure 5: ToF 3D Imaging of Optically Rough Surfaces using Band-Limited Synthetic Pulses: a) Setup schematic: The object (sphere) is illuminated by our band-limited CW tunable laser source and imaged with our holographic camera. b) The measured object, a spray-painted ball bearing with $50.4\ \mathrm{mm}$ diameter. c) Example synthetic pulse frame at $21\ \mathrm{ps}$. d) 3D reconstruction of the object surface from the SLiF measurement at $2.41\ \mathrm{nm}$ laser bandwidth. The obtained depth precision is $340\ \mu\mathrm{m}$, which is close to the theoretical axial resolution limit of $270\ \mu\mathrm{m}$. e) Second example for 3D surfaces reconstruction: Bunny surface, reconstructed from the measurements shown in Fig. \ref{['fig:Fig_1']}e using $0.23\ \mathrm{nm}$ of bandwidth.