Share Email Print
cover

Proceedings Paper

A photon-efficient method based on curve fitting for photon counting 3D imaging lidar
Author(s): Ling Ye; Guohua Gu; Weiji He; Wenye Yin; Jie Lin; Jian Fang
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Time-of-flight (TOF) Lidar is widely used in capturing three-dimensional (3D) structure and reflectivity information. For using Geiger-mode avalanche photodiode (Gm-APD) and the technique time correlated single photon counting (TCSPC), a direct-detection 3D imaging lidar has high sensitivity in low-light-level (LLL) scene. Traditional method needs long fixed dwell time to collect tens of thousands of photons to find accurate range and mitigate Poisson noise at each pixel. We present a method that acquires accurate depth and intensity images using a small amount of detected echo photons and having quantitative analysis to estimate whether results are in the confidence interval. Based on prior knowledge that the echo signal is in the shape of emitted laser, we use one or two orders of magnitude back-reflected photons less than traditional method, fitting a curve of laser-return pulse by nonlinear least-squares fitting in order to obtain the range. The condition of moving to next pixel in our method is acquiring a fixed number of back-reflected photons, instead of sampling for a fixed time. This adaptive jump condition is able to speed up the scanning without more distortion. The results are analyzed with chi-square test to determine if the curve we fit has enough credibility. This quantitative analysis provides an important judgment condition for our method of fitting curve to recover the depth image. Experimental results demonstrate that our method is able to obtain the millimeter accuracy depth image in the confidence interval using hundreds of photons and increases photon-efficiency more than 10-fold over traditional method. Thus our method will be useful in LLL scene, such as military reconnaissance and remote sensing.

Paper Details

Date Published: 18 September 2018
PDF: 6 pages
Proc. SPIE 10767, Remote Sensing and Modeling of Ecosystems for Sustainability XV, 107670X (18 September 2018); doi: 10.1117/12.2317805
Show Author Affiliations
Ling Ye, Nanjing Univ. of Sciences and Technology (China)
Guohua Gu, Nanjing Univ. of Sciences and Technology (China)
Weiji He, Nanjing Univ. of Sciences and Technology (China)
Wenye Yin, Nanjing Univ. of Sciences and Technology (China)
Jie Lin, Nanjing Univ. of Sciences and Technology (China)
Jian Fang, Nanjing Univ. of Sciences and Technology (China)


Published in SPIE Proceedings Vol. 10767:
Remote Sensing and Modeling of Ecosystems for Sustainability XV
Wei Gao; Ni-Bin Chang; Jinnian Wang, Editor(s)

© SPIE. Terms of Use
Back to Top