Augmenting text document by on-line learning of local arrangement of keypoints

Hideaki Uchiyama, Hideo Saito

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

15 Citations (Scopus)

Abstract

We propose a technique for text document tracking over a large range of viewpoints. Since the popular SIFT or SURF descriptors typically fail on such documents, our method considers instead local arrangement of keypoints. We extends Locally Likely Arrangement Hashing (LLAH), which is limited to fronto-parallel images: We handle a large range of viewpoints by learning the behavior of keypoint patterns when the camera viewpoint changes. Our method starts tracking a document from a nearly frontal view. Then, it undergoes motion, and new configurations of keypoints appear. The database is incrementally updated to reflect these new observations, allowing the system to detect the document under the new viewpoint. We demonstrate the performance and robustness of our method by comparing it with the original LLAH.

Original languageEnglish
Title of host publicationScience and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009
Pages95-98
Number of pages4
DOIs
Publication statusPublished - Dec 1 2009
Externally publishedYes
Event8th IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009 - Science and Technology - Orlando, FL, United States
Duration: Oct 19 2009Oct 22 2009

Publication series

NameScience and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009

Other

Other8th IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009 - Science and Technology
CountryUnited States
CityOrlando, FL
Period10/19/0910/22/09

Fingerprint

Cameras

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Computer Science Applications

Cite this

Uchiyama, H., & Saito, H. (2009). Augmenting text document by on-line learning of local arrangement of keypoints. In Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009 (pp. 95-98). [5336491] (Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009). https://doi.org/10.1109/ISMAR.2009.5336491

Augmenting text document by on-line learning of local arrangement of keypoints. / Uchiyama, Hideaki; Saito, Hideo.

Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009. 2009. p. 95-98 5336491 (Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009).

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

Uchiyama, H & Saito, H 2009, Augmenting text document by on-line learning of local arrangement of keypoints. in Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009., 5336491, Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009, pp. 95-98, 8th IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009 - Science and Technology, Orlando, FL, United States, 10/19/09. https://doi.org/10.1109/ISMAR.2009.5336491
Uchiyama H, Saito H. Augmenting text document by on-line learning of local arrangement of keypoints. In Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009. 2009. p. 95-98. 5336491. (Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009). https://doi.org/10.1109/ISMAR.2009.5336491
Uchiyama, Hideaki ; Saito, Hideo. / Augmenting text document by on-line learning of local arrangement of keypoints. Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009. 2009. pp. 95-98 (Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009).
@inproceedings{d8dbf00625bf4518ad1969561f419952,
title = "Augmenting text document by on-line learning of local arrangement of keypoints",
abstract = "We propose a technique for text document tracking over a large range of viewpoints. Since the popular SIFT or SURF descriptors typically fail on such documents, our method considers instead local arrangement of keypoints. We extends Locally Likely Arrangement Hashing (LLAH), which is limited to fronto-parallel images: We handle a large range of viewpoints by learning the behavior of keypoint patterns when the camera viewpoint changes. Our method starts tracking a document from a nearly frontal view. Then, it undergoes motion, and new configurations of keypoints appear. The database is incrementally updated to reflect these new observations, allowing the system to detect the document under the new viewpoint. We demonstrate the performance and robustness of our method by comparing it with the original LLAH.",
author = "Hideaki Uchiyama and Hideo Saito",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/ISMAR.2009.5336491",
language = "English",
isbn = "9781424453900",
series = "Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009",
pages = "95--98",
booktitle = "Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009",

}

TY - GEN

T1 - Augmenting text document by on-line learning of local arrangement of keypoints

AU - Uchiyama, Hideaki

AU - Saito, Hideo

PY - 2009/12/1

Y1 - 2009/12/1

N2 - We propose a technique for text document tracking over a large range of viewpoints. Since the popular SIFT or SURF descriptors typically fail on such documents, our method considers instead local arrangement of keypoints. We extends Locally Likely Arrangement Hashing (LLAH), which is limited to fronto-parallel images: We handle a large range of viewpoints by learning the behavior of keypoint patterns when the camera viewpoint changes. Our method starts tracking a document from a nearly frontal view. Then, it undergoes motion, and new configurations of keypoints appear. The database is incrementally updated to reflect these new observations, allowing the system to detect the document under the new viewpoint. We demonstrate the performance and robustness of our method by comparing it with the original LLAH.

AB - We propose a technique for text document tracking over a large range of viewpoints. Since the popular SIFT or SURF descriptors typically fail on such documents, our method considers instead local arrangement of keypoints. We extends Locally Likely Arrangement Hashing (LLAH), which is limited to fronto-parallel images: We handle a large range of viewpoints by learning the behavior of keypoint patterns when the camera viewpoint changes. Our method starts tracking a document from a nearly frontal view. Then, it undergoes motion, and new configurations of keypoints appear. The database is incrementally updated to reflect these new observations, allowing the system to detect the document under the new viewpoint. We demonstrate the performance and robustness of our method by comparing it with the original LLAH.

UR - http://www.scopus.com/inward/record.url?scp=74549124891&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=74549124891&partnerID=8YFLogxK

U2 - 10.1109/ISMAR.2009.5336491

DO - 10.1109/ISMAR.2009.5336491

M3 - Conference contribution

AN - SCOPUS:74549124891

SN - 9781424453900

T3 - Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009

SP - 95

EP - 98

BT - Science and Technology Proceedings - IEEE 2009 International Symposium on Mixed and Augmented Reality, ISMAR 2009

ER -