A Coding Theory of Time-of-Flight 3D Imaging

The depth resolution achieved by a continuous wave time-of-flight (C-ToF) imaging system is determined by the coding (modulation and demodulation) functions that it uses. Almost all current C ToF systems use sinusoid or square coding functions, resulting in a limited depth resolution. We present a mathematical framework for exploring and characterizing the space of C-ToF coding functions in a geometrically intuitive space. Using this framework, we design families of novel coding functions that are based on Hamiltonian cycles on hypercube graphs. Given a fixed power and acquisition time, the new Hamiltonian coding scheme can achieve up to an order of magnitude higher resolution as compared to the current state-of-the-art methods, especially in low signal-to-noise ratio (SNR) settings. Since most off-the-shelf C-ToF sensors use sinusoid or square functions, we develop a hardware prototype that can implement a wide range of coding functions. Using this prototype, we demonstrate the performance advantages of the proposed Hamiltonian coding functions in a wide range of imaging settings.

Publications

What Are Optimal Coding Functions for Time-of-Flight Imaging?

Mohit Gupta, S Nayar, A Velten, Eric Breitbach

ACM TOG, presented at SIGGRAPH 2018

Finalists in the Most Influential Research category AutoSens Awards, 2019

Practical Coding Function Design for Time-of-Flight Imaging

Felipe Gutierrez, SA Reza, A Velten, Mohit Gupta

Proc. CVPR 2019

A Geometric Perspective on Structured Light Coding

Mohit Gupta, Nikhil Nakhate

Proc. ECCV 2018

Video Results

Resources

Presentation Slides

Video

Comments

Share This Article

Share on LinkedInShare on FacebookTweet about this on Twitter