0
Technical Brief

Complete Sensitivity Analysis in a LiDAR-Camera Calibration Model

[+] Author and Article Information
Angel-Iván García-Moreno, Alfonso Ramírez-Pedraza, Juan B. Hurtado-Ramos, Francisco-Javier Ornelas-Rodríguez

Department of Image Analysis,
Research Center for Applied Science and Advanced Technology (CICATA),
Instituto Politécnico Nacional,
Cerro Blanco 141 Col. Colinas del Cimatario,
Quéretaro 76090, México

Jose-Joel González-Barbosa

Department of Image Analysis,
Research Center for Applied Science and Advanced Technology (CICATA),
Instituto Politécnico Nacional,
Cerro Blanco 141 Col. Colinas del Cimatario,
Quéretaro 76090, México
e-mail: jgonzalezba@ipn.mx

1Corresponding author.

Contributed by the Design Engineering Division of ASME for publication in the JOURNAL OF COMPUTING AND INFORMATION SCIENCE IN ENGINEERING. Manuscript received June 3, 2015; final manuscript received October 30, 2015; published online December 11, 2015. Editor: Bahram Ravani.

J. Comput. Inf. Sci. Eng 16(1), 014501 (Dec 11, 2015) (10 pages) Paper No: JCISE-15-1184; doi: 10.1115/1.4032026 History: Received June 03, 2015; Revised October 30, 2015

This paper presents a sensitivity analysis in the calibration of two sensors: a laser sensor Velodyne HDL-64E and a panoramic camera Point Grey Ladybug2; both sensors are used for three-dimensional urban reconstruction and were calibrated by two techniques; their results are compared in the sensitivity analysis. The effect of each parameter on the errors propagated in our platform reconstruction was analyzed using the simulated and experimental data sets. To compare the calibration parameters, we implement simulation techniques like Monte Carlo (MC) and Latin hypercube sampling (LHS). The sensitivity index for each parameter was calculated by two methods. The Sobol method, which is based on the analysis of the variance breakdown and the Fourier amplitude sensitivity test (FAST) method, based on Fourier analysis. Measures of variability on the simulated and experimental calibration parameters are shown for the calibration techniques implemented. Results on parameters and factors causing higher uncertainty in the calibration process are presented.

FIGURES IN THIS ARTICLE
<>
Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.

References

Saltelli, A. , Chan, K. , and Scott, E. M. , 2000, Sensitivity Analysis, Wiley, New York.
Helton, J. , Davis, F. , and Johnson, J. , 2005, “ A Comparison of Uncertainty and Sensitivity Analysis Results Obtained With Random and Latin Hypercube Sampling,” Reliab. Eng. Syst. Saf., 89(3), pp. 305–330. [CrossRef]
Cacuci, D. G. , Ionescu-Bujor, M. , and Navon, I. M. , 2005, Sensitivity and Uncertainty Analysis: Applications to Large-Scale systems, Vol. 2, CRC Press, Boca Raton, FL.
Saltelli, A. , Tarantola, S. , Campolongo, F. , and Ratto, M. , 2004, Sensitivity Analysis in Practice—A Guide to Assessing Scientific Models, Wiley, Hoboken, NJ.
Raymond, H. M. , 1999, “ Response Surface Methodology—Current Status and Future Directions,” J. Qual. Technol., 31(1), pp. 30–44, http://asq.org/qic/display-item/index.html?item=11563
Lilburne, L. , and Tarantola, S. , 2009, “ Sensitivity Analysis of Spatial Models,” Int. J. Geogr. Inf. Sci., 23(2), pp. 151–168. [CrossRef]
Martinelli, A. , 2012, “ Vision and IMU Data Fusion: Closed-Form Solutions for Attitude, Speed, Absolute Scale, and Bias Determination,” IEEE Trans. Rob., 28(1), pp. 44–60. [CrossRef]
Wijnholds, S. J. , and Van Der Veen, A.-J. , 2009, “ Multisource Self-Calibration for Sensor Arrays,” IEEE Trans. Sig. Process., 57(9), pp. 3512–3522. [CrossRef]
Doucet, A. , Godsill, S. , and Andrieu, C. , 2000, “ On Sequential Monte Carlo Sampling Methods for Bayesian Filtering,” Stat. Comput., 10(3), pp. 197–208. [CrossRef]
Pilger, G. G. , Costa, J. F. C. , and Koppe, J. C. , 2005, “ Improving the Efficiency of the Sequential Simulation Algorithm Using Latin Hypercube Sampling,” Geostatistics Banff 2004, Springer, The Netherlands.
Cropp, R. A. , and Braddock, R. D. , 2002, “ The New Morris Method: An Efficient Second-Order Screening Method,” Reliab. Eng. Syst. Saf., 78(1), pp. 77–83. [CrossRef]
Wallach, D. , Makowski, D. , and Jones, J. W. , 2006, Working With Dynamic Crop Models: Evaluation, Analysis, Parameterization, and Applications, Elsevier, Amsterdam.
Saltelli, A. , Annoni, P. , Azzini, I. , Campolongo, F. , Ratto, M. , and Tarantola, S. , 2010, “ Variance Based Sensitivity Analysis of Model Output. Design and Estimator for the Total Sensitivity Index,” Comput. Phys. Commun., 181(2), pp. 259–270. [CrossRef]
Atanacio-Jiménez, G. , González-Barbosa, J.-J. , Hurtado-Ramos, J. B. , Ornelas-Rodríguez, F. J. , Jiménez-Hernández, H. , García-Ramirez, T. , and González-Barbosa, R. , 2011, “ Lidar Velodyne HDL-64e Calibration Using Pattern Planes,” Int. J. Adv. Rob. Syst., 8(5), pp. 70–82.
Tsai, R. Y. , 1987, “ A Versatile Camera Calibration Technique for High-Accuracy 3D Machine Vision Metrology Using Off-The-Shelf Tv Cameras and Lenses,” IEEE J. Rob. Autom., 3(4), pp. 323–344. [CrossRef]
Zhang, Z. , 2000, “ A Flexible New Technique for Camera Calibration,” IEEE Trans. Pattern Anal. Mach. Intell., 22(11), pp. 1330–1334. [CrossRef]
García-Moreno, A.-I. , Gonzalez-Barbosa, J.-J. , Ornelas-Rodriguez, F.-J. , Hurtado-Ramos, J. B. , and Primo-Fuentes, M.-N. , 2013, “ Lidar and Panoramic Camera Extrinsic Calibration Approach Using a Pattern Plane,” Pattern Recognition, Springer, Berlin.
Guides in Metrology, 2008, Evaluation of Measurement Data—Guide to the Expression of Uncertainty in Measurement, Bureau International des Poids et Mesures, Geneève: International Organization for Standardization.
Archer, G. , Saltelli, A. , and Sobol, I. , 1997, “ Sensitivity Measures, Anova-Like Techniques and the Use of Bootstrap,” J. Stat. Comput. Simul., 58(2), pp. 99–120. [CrossRef]
Cannavó, F. , 2012, “ Sensitivity Analysis for Volcanic Source Modeling Quality Assessment and Model Selection,” Comput. Geosci., 44, pp. 52–59. [CrossRef]
Sen, S. K. , Samanta, T. , and Reese, A. , 2006, “ Quasi-Versus Pseudo-Random Generators: Discrepancy, Complexity and Integration-Error Based Comparison,” Int. J. Innovative Comput., Inf. Control, 2(3), pp. 621–651, http://www.ijicic.org/vol-2(3).htm
Sobol’, I. , 1967, “ On the Distribution of Points in a Cube and the Approximate Evaluation of Integrals,” {USSR} Comput. Math. Math. Phys., 7(4), pp. 86–112. [CrossRef]
Halton, J. H. , 1964, “ Algorithm 247: Radical-Inverse Quasi-Random Point Sequence,” Commun. ACM, 7(12), pp. 701–702. [CrossRef]
Helton, J. C. , and Davis, F. J. , 2003, “ Latin Hypercube Sampling and the Propagation of Uncertainty in Analyses of Complex Systems,” Reliab. Eng. Syst. Saf., 81(1), pp. 23–69. [CrossRef]
Conover, W. J. , and Iman, R. L. , 1981, “ Rank Transformations as a Bridge Between Parametric and Nonparametric Statistics,” Am. Stat., 35(3), pp. 124–129.
García-Moreno, A.-I. , Hernandez-García, D.-E. , Gonzalez-Barbosa, J.-J. , Ramírez-Pedraza, A. , Hurtado-Ramos, J. B. , and Ornelas-Rodriguez, F.-J. , 2014, “ Error Propagation and Uncertainty Analysis Between 3D Laser Scanner and Camera,” Rob. Auton. Syst., 62(6), pp. 782–793. [CrossRef]
Sun, W. , and Cooperstock, J. R. , 2006, “ An Empirical Evaluation of Factors Influencing Camera Calibration Accuracy Using Three Publicly Available Techniques,” Mach. Vision Appl., 17(1), pp. 51–67. [CrossRef]

Figures

Grahic Jump Location
Fig. 1

Sensor platform composed of LiDAR Velodyne HDL-64E (L) Ladybug2 (LB) and GPS. Ladybug2 spherical digital video camera system has six cameras (Ci). [R,T]LCi represents rotation and translation of the LiDAR and six camera frames of the Ladybug2.

Grahic Jump Location
Fig. 2

LiDAR framework L and the multicamera system C1,…,C5. [R,T]WCi represents the world-camera transformation, [R,T]WL represents the world-LiDAR transformation, and [R,T]LCi represents the LiDAR-camera transformation.

Grahic Jump Location
Fig. 3

LiDAR Velodyne HDL-64E configuration

Grahic Jump Location
Fig. 4

Tsai's calibration method analyzed by Sobol technique: blue bars are the global sensitivity (Si) and red bars are the total sensitivity (Sitot). Simulated data by MC and LHS with six different distributions. The most relevant parameters are tz and β, i.e., with more sensitive. Parameter identification is indicated in Table 2. (a) Sobol sequence by MC, (b) Halton sequence by MC, (c) normal distribution by MC, (d) normal distribution by LHS, (e) uniform distribution by LHS, and (f) multivariate distribution by LHS. (See online version for color.)

Grahic Jump Location
Fig. 5

Zhang's calibration method analyzed by Sobol technique: blue bars are the global sensitivity (Si) and red bars are the total sensitivity (Sitot). Simulated data by MC and LHS with six different distributions. The most relevant parameters are tz and β, i.e., with more sensitive. Parameter identification is indicated in Table 2. (a) Sobol sequence by MC, (b) Halton sequence by MC, (c) normal distribution by MC, (d) normal distribution by LHS, (e) uniform distribution by LHS, and (f) multivariate distribution by LHS. (See online version for color.)

Grahic Jump Location
Fig. 6

Experimental camera calibration results. Twenty sets of images were taken. For each set, our calibration algorithms were run 100 times. The left side of the graphs displays the average per parameter calibrated (dashed line). The right side shows the variability of these results (chain line). The greater the variability, the greater the sensitivity. The tz − ty and α−β parameters show greater change, similar results to those calculated by the FAST and Sobol techniques.

Grahic Jump Location
Fig. 7

LiDAR sensitivity analysis by Sobol method: blue bars are the global sensitivity (Si) and red bars are the total sensitivity (Sitot). Simulated data by MC and LHS with six different distributions. The most relevant parameters are θ and Δθ. Parameter identification is indicated in Table 3. (a) Sobol sequence by MC, (b) Halton sequence by MC, (c) normal distribution by MC, (d) normal distribution by LHS, (e) uniform distribution by LHS, and (f) multivariate distribution by LHS.

Grahic Jump Location
Fig. 8

Experimental LiDAR results. The left side of the graphs shows the average of each laser to calibrate 100 datasets (dashed line). The right side shows the variability of these results (chain line). The greater the variability, the greater the sensitivity. The Δvosc−Δθ parameters show greater change, similar results to those calculated by FAST and Sobol techniques.

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In