Towards a more accurate light transport model for non-line-of-sight imaging

Area of Science:

• Optics
• Computational Imaging
• Applied Physics

Background:

• Non-line-of-sight (NLOS) imaging reconstructs hidden scenes using optical signals reflected off diffuse surfaces.
• Existing NLOS techniques include photon time-of-flight (ToF) and shadow-based methods, each with separate forward models.
• A unified framework, the two-frequency spatial Wigner distribution (TFSWD), was previously proposed but limited to known occluder geometry.

Purpose of the Study:

• To develop a graphical representation of the TFSWD forward model for NLOS imaging.
• To apply this unified framework to novel experimental setups.
• To explore combining ToF-NLOS and Shadow-NLOS modalities when occluder geometry is unknown.

Main Methods:

• Developed a graphical representation of the two-frequency spatial Wigner distribution (TFSWD) forward model.
• Applied the TFSWD model to new experimental configurations for NLOS imaging.
• Investigated the combination of ToF-NLOS and Shadow-NLOS imaging within the unified TFSWD framework.

Main Results:

• Presented a novel graphical interpretation of the TFSWD forward model.
• Demonstrated the application of the unified model in experimental NLOS imaging scenarios.
• Showcased the potential for combining ToF and shadow-based NLOS imaging without prior occluder information.

Conclusions:

• The unified TFSWD framework offers a versatile approach to NLOS imaging.
• Graphical representation simplifies understanding and application of the TFSWD model.
• This work advances the combination of different NLOS imaging modalities for improved scene reconstruction, particularly in complex scenarios.