Bayesian 3D Reconstruction of Complex Scenes from Single-Photon Lidar Data

Julián Tachella, Yoann Altmann, Ximing Ren, Aongus McCarthy, Gerald Stuart Buller, Stephen McLaughlin, Jean-Yves Tourneret

Research output: Contribution to journalArticle

Abstract

Light detection and ranging (Lidar) data can be used to capture the depth and intensity profile of a 3D scene. This modality relies on constructing, for each pixel, a histogram of time delays between emitted light pulses and detected photon arrivals. In a general setting, more than one surface can be observed in a single pixel. The problem of estimating the number of surfaces, their reflectivity and position becomes very challenging in the low-photon regime (which equates to short acquisition times) or relatively high background levels (i.e., strong ambient illumination). This paper presents a new approach to 3D reconstruction using single-photon, single-wavelength Lidar data, which is capable of identifying multiple surfaces in each pixel. Adopting a Bayesian approach, the 3D structure to be recovered is modelled as a marked point process and reversible jump Markov chain Monte Carlo (RJ-MCMC) moves are proposed to sample the posterior distribution of interest. In order to promote spatial correlation between points belonging to the same surface, we propose a prior that combines an area interaction process and a Strauss process. New RJ-MCMC dilation and erosion updates are presented to achieve an efficient exploration of the configuration space. To further reduce the computational load, we adopt a multiresolution approach, processing the data from a coarse to the finest scale. The experiments performed with synthetic and real data show that the algorithm obtains better reconstructions than other recently published optimization algorithms for lower execution times.
LanguageEnglish
Pages521-550
Number of pages30
JournalSIAM Journal on Imaging Sciences
Volume12
Issue number1
DOIs
Publication statusPublished - 14 Mar 2019

Fingerprint

Photons
Pixels
Markov processes
Erosion
Time delay
Lighting
Wavelength
Processing
Experiments

Keywords

  • 3D reconstruction
  • Bayesian inference
  • Lidar
  • Low-photon imaging
  • Poisson noise

ASJC Scopus subject areas

  • Mathematics(all)
  • Applied Mathematics

Cite this

@article{b11027eb792e4dcf98096ebc5c8ef138,
title = "Bayesian 3D Reconstruction of Complex Scenes from Single-Photon Lidar Data",
abstract = "Light detection and ranging (Lidar) data can be used to capture the depth and intensity profile of a 3D scene. This modality relies on constructing, for each pixel, a histogram of time delays between emitted light pulses and detected photon arrivals. In a general setting, more than one surface can be observed in a single pixel. The problem of estimating the number of surfaces, their reflectivity and position becomes very challenging in the low-photon regime (which equates to short acquisition times) or relatively high background levels (i.e., strong ambient illumination). This paper presents a new approach to 3D reconstruction using single-photon, single-wavelength Lidar data, which is capable of identifying multiple surfaces in each pixel. Adopting a Bayesian approach, the 3D structure to be recovered is modelled as a marked point process and reversible jump Markov chain Monte Carlo (RJ-MCMC) moves are proposed to sample the posterior distribution of interest. In order to promote spatial correlation between points belonging to the same surface, we propose a prior that combines an area interaction process and a Strauss process. New RJ-MCMC dilation and erosion updates are presented to achieve an efficient exploration of the configuration space. To further reduce the computational load, we adopt a multiresolution approach, processing the data from a coarse to the finest scale. The experiments performed with synthetic and real data show that the algorithm obtains better reconstructions than other recently published optimization algorithms for lower execution times.",
keywords = "3D reconstruction, Bayesian inference, Lidar, Low-photon imaging, Poisson noise",
author = "Juli{\'a}n Tachella and Yoann Altmann and Ximing Ren and Aongus McCarthy and Buller, {Gerald Stuart} and Stephen McLaughlin and Jean-Yves Tourneret",
year = "2019",
month = "3",
day = "14",
doi = "10.1137/18M1183972",
language = "English",
volume = "12",
pages = "521--550",
journal = "SIAM Journal on Imaging Sciences",
issn = "1936-4954",
publisher = "Society for Industrial and Applied Mathematics Publications",
number = "1",

}

Bayesian 3D Reconstruction of Complex Scenes from Single-Photon Lidar Data. / Tachella, Julián; Altmann, Yoann; Ren, Ximing; McCarthy, Aongus; Buller, Gerald Stuart; McLaughlin, Stephen; Tourneret, Jean-Yves.

In: SIAM Journal on Imaging Sciences , Vol. 12, No. 1, 14.03.2019, p. 521-550.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Bayesian 3D Reconstruction of Complex Scenes from Single-Photon Lidar Data

AU - Tachella, Julián

AU - Altmann, Yoann

AU - Ren, Ximing

AU - McCarthy, Aongus

AU - Buller, Gerald Stuart

AU - McLaughlin, Stephen

AU - Tourneret, Jean-Yves

PY - 2019/3/14

Y1 - 2019/3/14

N2 - Light detection and ranging (Lidar) data can be used to capture the depth and intensity profile of a 3D scene. This modality relies on constructing, for each pixel, a histogram of time delays between emitted light pulses and detected photon arrivals. In a general setting, more than one surface can be observed in a single pixel. The problem of estimating the number of surfaces, their reflectivity and position becomes very challenging in the low-photon regime (which equates to short acquisition times) or relatively high background levels (i.e., strong ambient illumination). This paper presents a new approach to 3D reconstruction using single-photon, single-wavelength Lidar data, which is capable of identifying multiple surfaces in each pixel. Adopting a Bayesian approach, the 3D structure to be recovered is modelled as a marked point process and reversible jump Markov chain Monte Carlo (RJ-MCMC) moves are proposed to sample the posterior distribution of interest. In order to promote spatial correlation between points belonging to the same surface, we propose a prior that combines an area interaction process and a Strauss process. New RJ-MCMC dilation and erosion updates are presented to achieve an efficient exploration of the configuration space. To further reduce the computational load, we adopt a multiresolution approach, processing the data from a coarse to the finest scale. The experiments performed with synthetic and real data show that the algorithm obtains better reconstructions than other recently published optimization algorithms for lower execution times.

AB - Light detection and ranging (Lidar) data can be used to capture the depth and intensity profile of a 3D scene. This modality relies on constructing, for each pixel, a histogram of time delays between emitted light pulses and detected photon arrivals. In a general setting, more than one surface can be observed in a single pixel. The problem of estimating the number of surfaces, their reflectivity and position becomes very challenging in the low-photon regime (which equates to short acquisition times) or relatively high background levels (i.e., strong ambient illumination). This paper presents a new approach to 3D reconstruction using single-photon, single-wavelength Lidar data, which is capable of identifying multiple surfaces in each pixel. Adopting a Bayesian approach, the 3D structure to be recovered is modelled as a marked point process and reversible jump Markov chain Monte Carlo (RJ-MCMC) moves are proposed to sample the posterior distribution of interest. In order to promote spatial correlation between points belonging to the same surface, we propose a prior that combines an area interaction process and a Strauss process. New RJ-MCMC dilation and erosion updates are presented to achieve an efficient exploration of the configuration space. To further reduce the computational load, we adopt a multiresolution approach, processing the data from a coarse to the finest scale. The experiments performed with synthetic and real data show that the algorithm obtains better reconstructions than other recently published optimization algorithms for lower execution times.

KW - 3D reconstruction

KW - Bayesian inference

KW - Lidar

KW - Low-photon imaging

KW - Poisson noise

UR - http://www.scopus.com/inward/record.url?scp=85064214850&partnerID=8YFLogxK

U2 - 10.1137/18M1183972

DO - 10.1137/18M1183972

M3 - Article

VL - 12

SP - 521

EP - 550

JO - SIAM Journal on Imaging Sciences

T2 - SIAM Journal on Imaging Sciences

JF - SIAM Journal on Imaging Sciences

SN - 1936-4954

IS - 1

ER -