Related Experiment Video
Updated: Sep 20, 2025

11:57
Measuring Spatially- and Directionally-varying Light Scattering from Biological Material
Published on: May 20, 2013
13.6K
A boundary migration model for imaging within volumetric scattering media.
Dongyu Du1, Xin Jin2, Rujia Deng1
1Shenzhen International Graduate School, Tsinghua University, 518055, Shenzhen, China.
Nature Communications
|June 10, 2022
Summary
This study introduces a novel boundary migration model (BMM) for imaging Lambertian objects in highly scattering media. The BMM effectively reconstructs objects even with minimal signal photons, outperforming existing methods.
Area of Science:
- Optics
- Imaging Science
- Computational Imaging
Background:
- Imaging through scattering media is challenging due to signal attenuation and background noise.
- Existing methods struggle with highly scattering environments and low signal-to-noise ratios.
- Macroscopic imaging in scattering media requires robust techniques.
Purpose of the Study:
- To develop a novel method for imaging Lambertian objects embedded in highly scattering media.
- To address challenges of signal attenuation and background photon coupling.
- To improve reconstruction precision and scattering strength compared to existing techniques.
Main Methods:
- A time-to-space boundary migration model (BMM) was developed to process scattered optical signals.
- The BMM converts spectral measurements into temporal domain scene information.
- Experiments were conducted on 2D and 3D Lambertian objects in polyethylene foam and fog.
Main Results:
- The proposed BMM successfully reconstructed Lambertian objects in highly scattering media.
- Reconstruction was achieved even with only 0.75% signal photons at depths exceeding 25 transport mean free paths.
- The method demonstrated superior reconstruction precision and scattering strength over time gating.
Conclusions:
- The boundary migration model (BMM) offers an effective solution for imaging in challenging scattering conditions.
- The technique shows significant potential for macroscopic imaging applications due to its precision and speed.
- Low reconstruction complexity and millisecond-scale runtime facilitate practical implementation.

