Human-readable fiducial marker classification using convolutional neural networks

Yanfeng Liu, Eric T. Psota, Lance C Perez

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Many applications require both the location and identity of objects in images and video. Most existing solutions, like QR codes, AprilTags, and ARTags use complex machine-readable fiducial markers with heuristically derived methods for detection and classification. However, in applications where humans are integral to the system and need to be capable of locating objects in the environment, fiducial markers must be human readable. An obvious and convenient choice for human readable fiducial markers are alphanumeric characters (Arabic numbers and English letters). Here, a method for classifying characters using a convolutional neural network (CNN) is presented. The network is trained with a large set of computer generated images of characters where each is subjected to a carefully designed set of augmentations designed to simulate the conditions inherent in video capture. These augmentations include rotation, scaling, shearing, and blur. Results demonstrate that training on large numbers of synthetic images produces a system that works on real images captured by a video camera. The result also reveal that certain characters are generally more reliable and easier to recognize than others, thus the results can be used to intelligently design a human-readable fiducial markers system that avoids confusing characters.

Original languageEnglish (US)
Title of host publication2017 IEEE International Conference on Electro Information Technology, EIT 2017
PublisherIEEE Computer Society
Pages606-610
Number of pages5
ISBN (Electronic)9781509047673
DOIs
StatePublished - Sep 27 2017
Event2017 IEEE International Conference on Electro Information Technology, EIT 2017 - Lincoln, United States
Duration: May 14 2017May 17 2017

Publication series

NameIEEE International Conference on Electro Information Technology
ISSN (Print)2154-0357
ISSN (Electronic)2154-0373

Other

Other2017 IEEE International Conference on Electro Information Technology, EIT 2017
CountryUnited States
CityLincoln
Period5/14/175/17/17

Fingerprint

Neural networks
Video cameras
Shearing

Keywords

  • computer vision
  • convolutional neural network
  • fiducial marker
  • machine learning

ASJC Scopus subject areas

  • Computer Science Applications
  • Information Systems
  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Cite this

Liu, Y., Psota, E. T., & Perez, L. C. (2017). Human-readable fiducial marker classification using convolutional neural networks. In 2017 IEEE International Conference on Electro Information Technology, EIT 2017 (pp. 606-610). [8053435] (IEEE International Conference on Electro Information Technology). IEEE Computer Society. https://doi.org/10.1109/EIT.2017.8053435

Human-readable fiducial marker classification using convolutional neural networks. / Liu, Yanfeng; Psota, Eric T.; Perez, Lance C.

2017 IEEE International Conference on Electro Information Technology, EIT 2017. IEEE Computer Society, 2017. p. 606-610 8053435 (IEEE International Conference on Electro Information Technology).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Liu, Y, Psota, ET & Perez, LC 2017, Human-readable fiducial marker classification using convolutional neural networks. in 2017 IEEE International Conference on Electro Information Technology, EIT 2017., 8053435, IEEE International Conference on Electro Information Technology, IEEE Computer Society, pp. 606-610, 2017 IEEE International Conference on Electro Information Technology, EIT 2017, Lincoln, United States, 5/14/17. https://doi.org/10.1109/EIT.2017.8053435
Liu Y, Psota ET, Perez LC. Human-readable fiducial marker classification using convolutional neural networks. In 2017 IEEE International Conference on Electro Information Technology, EIT 2017. IEEE Computer Society. 2017. p. 606-610. 8053435. (IEEE International Conference on Electro Information Technology). https://doi.org/10.1109/EIT.2017.8053435
Liu, Yanfeng ; Psota, Eric T. ; Perez, Lance C. / Human-readable fiducial marker classification using convolutional neural networks. 2017 IEEE International Conference on Electro Information Technology, EIT 2017. IEEE Computer Society, 2017. pp. 606-610 (IEEE International Conference on Electro Information Technology).
@inproceedings{459614798c4745bd93b275980b814a84,
title = "Human-readable fiducial marker classification using convolutional neural networks",
abstract = "Many applications require both the location and identity of objects in images and video. Most existing solutions, like QR codes, AprilTags, and ARTags use complex machine-readable fiducial markers with heuristically derived methods for detection and classification. However, in applications where humans are integral to the system and need to be capable of locating objects in the environment, fiducial markers must be human readable. An obvious and convenient choice for human readable fiducial markers are alphanumeric characters (Arabic numbers and English letters). Here, a method for classifying characters using a convolutional neural network (CNN) is presented. The network is trained with a large set of computer generated images of characters where each is subjected to a carefully designed set of augmentations designed to simulate the conditions inherent in video capture. These augmentations include rotation, scaling, shearing, and blur. Results demonstrate that training on large numbers of synthetic images produces a system that works on real images captured by a video camera. The result also reveal that certain characters are generally more reliable and easier to recognize than others, thus the results can be used to intelligently design a human-readable fiducial markers system that avoids confusing characters.",
keywords = "computer vision, convolutional neural network, fiducial marker, machine learning",
author = "Yanfeng Liu and Psota, {Eric T.} and Perez, {Lance C}",
year = "2017",
month = "9",
day = "27",
doi = "10.1109/EIT.2017.8053435",
language = "English (US)",
series = "IEEE International Conference on Electro Information Technology",
publisher = "IEEE Computer Society",
pages = "606--610",
booktitle = "2017 IEEE International Conference on Electro Information Technology, EIT 2017",

}

TY - GEN

T1 - Human-readable fiducial marker classification using convolutional neural networks

AU - Liu, Yanfeng

AU - Psota, Eric T.

AU - Perez, Lance C

PY - 2017/9/27

Y1 - 2017/9/27

N2 - Many applications require both the location and identity of objects in images and video. Most existing solutions, like QR codes, AprilTags, and ARTags use complex machine-readable fiducial markers with heuristically derived methods for detection and classification. However, in applications where humans are integral to the system and need to be capable of locating objects in the environment, fiducial markers must be human readable. An obvious and convenient choice for human readable fiducial markers are alphanumeric characters (Arabic numbers and English letters). Here, a method for classifying characters using a convolutional neural network (CNN) is presented. The network is trained with a large set of computer generated images of characters where each is subjected to a carefully designed set of augmentations designed to simulate the conditions inherent in video capture. These augmentations include rotation, scaling, shearing, and blur. Results demonstrate that training on large numbers of synthetic images produces a system that works on real images captured by a video camera. The result also reveal that certain characters are generally more reliable and easier to recognize than others, thus the results can be used to intelligently design a human-readable fiducial markers system that avoids confusing characters.

AB - Many applications require both the location and identity of objects in images and video. Most existing solutions, like QR codes, AprilTags, and ARTags use complex machine-readable fiducial markers with heuristically derived methods for detection and classification. However, in applications where humans are integral to the system and need to be capable of locating objects in the environment, fiducial markers must be human readable. An obvious and convenient choice for human readable fiducial markers are alphanumeric characters (Arabic numbers and English letters). Here, a method for classifying characters using a convolutional neural network (CNN) is presented. The network is trained with a large set of computer generated images of characters where each is subjected to a carefully designed set of augmentations designed to simulate the conditions inherent in video capture. These augmentations include rotation, scaling, shearing, and blur. Results demonstrate that training on large numbers of synthetic images produces a system that works on real images captured by a video camera. The result also reveal that certain characters are generally more reliable and easier to recognize than others, thus the results can be used to intelligently design a human-readable fiducial markers system that avoids confusing characters.

KW - computer vision

KW - convolutional neural network

KW - fiducial marker

KW - machine learning

UR - http://www.scopus.com/inward/record.url?scp=85033700518&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85033700518&partnerID=8YFLogxK

U2 - 10.1109/EIT.2017.8053435

DO - 10.1109/EIT.2017.8053435

M3 - Conference contribution

AN - SCOPUS:85033700518

T3 - IEEE International Conference on Electro Information Technology

SP - 606

EP - 610

BT - 2017 IEEE International Conference on Electro Information Technology, EIT 2017

PB - IEEE Computer Society

ER -