Quantifying error introduced by iterative closest point image registration

Ningjia Sun; Thomas Bull; Rupert Austin; David Bartlett; Saoirse O'Toole

doi:10.1016/j.jdent.2024.104863

Quantifying error introduced by iterative closest point image registration

Ningjia Sun^*, Thomas Bull, Rupert Austin, David Bartlett, Saoirse O'Toole

^*Corresponding author for this work

University of Southampton

Research output: Contribution to journal › Article › peer-review

41 Downloads (Pure)

Abstract

Objectives: The aim of this paper was to quantify the analysis error introduced by iterative closest point (ICP) image registration. We also investigated whether a subsequent subtraction process can reduce process error. Methods: We tested metrology and two 3D inspection software using calibration standards at 0.39 μm, and 2.64 μm and mathematically perfect defects (softgauges) at 2 and 20 μm, on free form surfaces of increasing complexity and area, both with and without registration. Errors were calculated in percentage relative to the size of the defect being measured. Data were analysed in GraphPad Prism 9, normal and two-way ANOVA with post-hoc Tukey's was applied. Significance was inferred at p < 0.05. Results: Using ICP registration introduced errors from 0 % to 15.63 % of the defect size depending on the surface complexity and size of the defect. Significant differences were observed in analysis measurements between metrology and 3D inspection software and within different 3D inspection software, however, one did not show clear superiority over another. Even in the absence of registration, defects at 0.39 μm, and 2.64 μm produced substantial measurement error (13.39–77.50 % of defect size) when using 3D inspection software. Adding an additional data subtraction process reduced registration error to negligible levels (<1 % independent of surface complexity or area). Conclusions: Commercial 3D inspection software introduces error during direct measurements below 3 μm. When using an ICP registration, errors over 15 % of the defect size can be introduced regardless of the accuracy of adjacent registration surfaces. Analysis output between software are not consistently repeatable or comparable and do not utilise ISO standards. Subtracting the datasets and analysing the residual difference reduced error to negligible levels. Clinical significance: This paper quantifies the significant errors and inconsistencies introduced during the registration process even when 3D datasets are true and precise. This may impact on research diagnostics and clinical performance. An additional data processing step of scan subtraction can reduce this error but increases computational complexity.

Original language	English
Article number	104863
Journal	Journal of Dentistry
Volume	142
Early online date	26 Jan 2024
DOIs	https://doi.org/10.1016/j.jdent.2024.104863
Publication status	Published - Mar 2024

Keywords

Dental informatics/bioinformatics
Diagnostic systems
Dimensional change
Imaging
Oral diagnosis
Surface metrology software

Access to Document

10.1016/j.jdent.2024.104863Licence: CC BY

Quantifying error introduced by_SUN_Publishedonline26January2024_GOLD VoR (CC BY)Final published version, 1.89 MBLicence: CC BY

Cite this

@article{9e87109eef2a43f2a12941bab645b3a6,

title = "Quantifying error introduced by iterative closest point image registration",

abstract = "Objectives: The aim of this paper was to quantify the analysis error introduced by iterative closest point (ICP) image registration. We also investigated whether a subsequent subtraction process can reduce process error. Methods: We tested metrology and two 3D inspection software using calibration standards at 0.39 μm, and 2.64 μm and mathematically perfect defects (softgauges) at 2 and 20 μm, on free form surfaces of increasing complexity and area, both with and without registration. Errors were calculated in percentage relative to the size of the defect being measured. Data were analysed in GraphPad Prism 9, normal and two-way ANOVA with post-hoc Tukey's was applied. Significance was inferred at p < 0.05. Results: Using ICP registration introduced errors from 0 % to 15.63 % of the defect size depending on the surface complexity and size of the defect. Significant differences were observed in analysis measurements between metrology and 3D inspection software and within different 3D inspection software, however, one did not show clear superiority over another. Even in the absence of registration, defects at 0.39 μm, and 2.64 μm produced substantial measurement error (13.39–77.50 % of defect size) when using 3D inspection software. Adding an additional data subtraction process reduced registration error to negligible levels (<1 % independent of surface complexity or area). Conclusions: Commercial 3D inspection software introduces error during direct measurements below 3 μm. When using an ICP registration, errors over 15 % of the defect size can be introduced regardless of the accuracy of adjacent registration surfaces. Analysis output between software are not consistently repeatable or comparable and do not utilise ISO standards. Subtracting the datasets and analysing the residual difference reduced error to negligible levels. Clinical significance: This paper quantifies the significant errors and inconsistencies introduced during the registration process even when 3D datasets are true and precise. This may impact on research diagnostics and clinical performance. An additional data processing step of scan subtraction can reduce this error but increases computational complexity.",

keywords = "Dental informatics/bioinformatics, Diagnostic systems, Dimensional change, Imaging, Oral diagnosis, Surface metrology software",

author = "Ningjia Sun and Thomas Bull and Rupert Austin and David Bartlett and Saoirse O'Toole",

note = "Funding Information: This work was supported by the King's-China Scholarship Council [grant number 202106230065]. Funding Information: This work was supported by the King's- China Scholarship Council [grant number 202106230065 ]. Publisher Copyright: {\textcopyright} 2024",

year = "2024",

month = mar,

doi = "10.1016/j.jdent.2024.104863",

language = "English",

volume = "142",

journal = "Journal of Dentistry",

issn = "0300-5712",

publisher = "Elsevier Ltd",

}

TY - JOUR

T1 - Quantifying error introduced by iterative closest point image registration

AU - Sun, Ningjia

AU - Bull, Thomas

AU - Austin, Rupert

AU - Bartlett, David

AU - O'Toole, Saoirse

N1 - Funding Information: This work was supported by the King's-China Scholarship Council [grant number 202106230065]. Funding Information: This work was supported by the King's- China Scholarship Council [grant number 202106230065 ]. Publisher Copyright: © 2024

PY - 2024/3

Y1 - 2024/3

N2 - Objectives: The aim of this paper was to quantify the analysis error introduced by iterative closest point (ICP) image registration. We also investigated whether a subsequent subtraction process can reduce process error. Methods: We tested metrology and two 3D inspection software using calibration standards at 0.39 μm, and 2.64 μm and mathematically perfect defects (softgauges) at 2 and 20 μm, on free form surfaces of increasing complexity and area, both with and without registration. Errors were calculated in percentage relative to the size of the defect being measured. Data were analysed in GraphPad Prism 9, normal and two-way ANOVA with post-hoc Tukey's was applied. Significance was inferred at p < 0.05. Results: Using ICP registration introduced errors from 0 % to 15.63 % of the defect size depending on the surface complexity and size of the defect. Significant differences were observed in analysis measurements between metrology and 3D inspection software and within different 3D inspection software, however, one did not show clear superiority over another. Even in the absence of registration, defects at 0.39 μm, and 2.64 μm produced substantial measurement error (13.39–77.50 % of defect size) when using 3D inspection software. Adding an additional data subtraction process reduced registration error to negligible levels (<1 % independent of surface complexity or area). Conclusions: Commercial 3D inspection software introduces error during direct measurements below 3 μm. When using an ICP registration, errors over 15 % of the defect size can be introduced regardless of the accuracy of adjacent registration surfaces. Analysis output between software are not consistently repeatable or comparable and do not utilise ISO standards. Subtracting the datasets and analysing the residual difference reduced error to negligible levels. Clinical significance: This paper quantifies the significant errors and inconsistencies introduced during the registration process even when 3D datasets are true and precise. This may impact on research diagnostics and clinical performance. An additional data processing step of scan subtraction can reduce this error but increases computational complexity.

AB - Objectives: The aim of this paper was to quantify the analysis error introduced by iterative closest point (ICP) image registration. We also investigated whether a subsequent subtraction process can reduce process error. Methods: We tested metrology and two 3D inspection software using calibration standards at 0.39 μm, and 2.64 μm and mathematically perfect defects (softgauges) at 2 and 20 μm, on free form surfaces of increasing complexity and area, both with and without registration. Errors were calculated in percentage relative to the size of the defect being measured. Data were analysed in GraphPad Prism 9, normal and two-way ANOVA with post-hoc Tukey's was applied. Significance was inferred at p < 0.05. Results: Using ICP registration introduced errors from 0 % to 15.63 % of the defect size depending on the surface complexity and size of the defect. Significant differences were observed in analysis measurements between metrology and 3D inspection software and within different 3D inspection software, however, one did not show clear superiority over another. Even in the absence of registration, defects at 0.39 μm, and 2.64 μm produced substantial measurement error (13.39–77.50 % of defect size) when using 3D inspection software. Adding an additional data subtraction process reduced registration error to negligible levels (<1 % independent of surface complexity or area). Conclusions: Commercial 3D inspection software introduces error during direct measurements below 3 μm. When using an ICP registration, errors over 15 % of the defect size can be introduced regardless of the accuracy of adjacent registration surfaces. Analysis output between software are not consistently repeatable or comparable and do not utilise ISO standards. Subtracting the datasets and analysing the residual difference reduced error to negligible levels. Clinical significance: This paper quantifies the significant errors and inconsistencies introduced during the registration process even when 3D datasets are true and precise. This may impact on research diagnostics and clinical performance. An additional data processing step of scan subtraction can reduce this error but increases computational complexity.

KW - Dental informatics/bioinformatics

KW - Diagnostic systems

KW - Dimensional change

KW - Imaging

KW - Oral diagnosis

KW - Surface metrology software

UR - http://www.scopus.com/inward/record.url?scp=85183487683&partnerID=8YFLogxK

U2 - 10.1016/j.jdent.2024.104863

DO - 10.1016/j.jdent.2024.104863

M3 - Article

SN - 0300-5712

VL - 142

JO - Journal of Dentistry

JF - Journal of Dentistry

M1 - 104863

ER -

Quantifying error introduced by iterative closest point image registration

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this