ConvNet-Based Optical Recognition for Engineering Drawings

Andrew Brock; Theodore Lim; James Millar Ritchie; Nicholas Weston

doi:10.1115/DETC2017-68186

ConvNet-Based Optical Recognition for Engineering Drawings

Andrew Brock, Theodore Lim, James Millar Ritchie, Nicholas Weston

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

2 Citations (Scopus)

Abstract

End-to-end machine analysis of engineering document drawings requires a reliable and precise vision frontend capable of localizing and classifying various characters in context. We develop an object detection framework, based on convolutional networks, designed specifically for optical character recognition in engineering drawings. Our approach enables classification and localization on a 10-fold cross-validation of an internal dataset for which other techniques prove unsuitable.

Original language	English
Title of host publication	ASME 2017 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference
Publisher	American Society of Mechanical Engineers
Volume	1
ISBN (Print)	9780791858110
DOIs	https://doi.org/10.1115/DETC2017-68186
Publication status	Published - 9 Aug 2017
Event	37th Computers and Information in Engineering Conference - Cleveland, United States Duration: 6 Aug 2017 → 9 Aug 2017

Conference

Conference	37th Computers and Information in Engineering Conference
Country/Territory	United States
City	Cleveland
Period	6/08/17 → 9/08/17

Keywords

Covnet
Optical recognition
GD&T
Feature recognition

Access to Document

10.1115/DETC2017-68186

Cite this

@inproceedings{02bd0caf7ce243a1b235e3fd7a02eaf3,

title = "ConvNet-Based Optical Recognition for Engineering Drawings",

abstract = "End-to-end machine analysis of engineering document drawings requires a reliable and precise vision frontend capable of localizing and classifying various characters in context. We develop an object detection framework, based on convolutional networks, designed specifically for optical character recognition in engineering drawings. Our approach enables classification and localization on a 10-fold cross-validation of an internal dataset for which other techniques prove unsuitable.",

keywords = "Covnet, Optical recognition, GD&T, Feature recognition",

author = "Andrew Brock and Theodore Lim and Ritchie, {James Millar} and Nicholas Weston",

year = "2017",

month = aug,

day = "9",

doi = "10.1115/DETC2017-68186",

language = "English",

isbn = "9780791858110",

volume = "1",

booktitle = "ASME 2017 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference",

publisher = "American Society of Mechanical Engineers",

address = "United States",

note = "37th Computers and Information in Engineering Conference ; Conference date: 06-08-2017 Through 09-08-2017",

}

Brock, A, Lim, T, Ritchie, JM & Weston, N 2017, ConvNet-Based Optical Recognition for Engineering Drawings. in ASME 2017 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. vol. 1, DETC2017-68186, American Society of Mechanical Engineers, 37th Computers and Information in Engineering Conference, Cleveland, Ohio, United States, 6/08/17. https://doi.org/10.1115/DETC2017-68186

ConvNet-Based Optical Recognition for Engineering Drawings. / Brock, Andrew; Lim, Theodore; Ritchie, James Millar et al.
ASME 2017 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference. Vol. 1 American Society of Mechanical Engineers, 2017. DETC2017-68186.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - ConvNet-Based Optical Recognition for Engineering Drawings

AU - Brock, Andrew

AU - Lim, Theodore

AU - Ritchie, James Millar

AU - Weston, Nicholas

PY - 2017/8/9

Y1 - 2017/8/9

N2 - End-to-end machine analysis of engineering document drawings requires a reliable and precise vision frontend capable of localizing and classifying various characters in context. We develop an object detection framework, based on convolutional networks, designed specifically for optical character recognition in engineering drawings. Our approach enables classification and localization on a 10-fold cross-validation of an internal dataset for which other techniques prove unsuitable.

AB - End-to-end machine analysis of engineering document drawings requires a reliable and precise vision frontend capable of localizing and classifying various characters in context. We develop an object detection framework, based on convolutional networks, designed specifically for optical character recognition in engineering drawings. Our approach enables classification and localization on a 10-fold cross-validation of an internal dataset for which other techniques prove unsuitable.

KW - Covnet

KW - Optical recognition

KW - GD&T

KW - Feature recognition

U2 - 10.1115/DETC2017-68186

DO - 10.1115/DETC2017-68186

M3 - Conference contribution

SN - 9780791858110

VL - 1

BT - ASME 2017 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference

PB - American Society of Mechanical Engineers

T2 - 37th Computers and Information in Engineering Conference

Y2 - 6 August 2017 through 9 August 2017

ER -

ConvNet-Based Optical Recognition for Engineering Drawings

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this