Hardware-aware Coding Function Design for Compressive Single-photon 3D Cameras

Area of Science:

• Optics and Photonics
• Computer Vision
• Signal Processing

Background:

• Single-photon cameras are crucial for high-resolution time-of-flight 3D imaging.
• Hardware limitations (bandwidth, laser power, data rates) hinder performance.
• Compressive histograms offer in-sensor data compression but underperform under hardware constraints.

Purpose of the Study:

• To develop practical coding functions for compressive single-photon 3D imaging.
• To address underperformance of compressive histograms under real-world hardware constraints.
• To optimize illumination and coding matrices within hardware limitations.

Main Methods:

• A constrained optimization approach using gradient descent.
• Joint optimization of illumination and coding matrices.
• Simulation-based evaluation and real-world system testing.

Main Results:

• Optimized coding functions outperform traditional designs under bandwidth and peak power constraints.
• Significant performance gains observed in peak power-constrained systems.
• The approach adapts to non-ideal impulse response functions.

Conclusions:

• The proposed constrained optimization method enhances compressive single-photon 3D imaging.
• This approach effectively overcomes hardware limitations in practical imaging systems.
• It offers a robust solution for real-world 3D imaging applications.