Abstract
In this paper, a new method is presented to extract both superimposed and embedded scene texts in digital news videos. The algorithm is summarized in the following three steps: preprocessing, extracting candidate regions, and filtering candidate regions. For the first preprocessing step, a color image is converted into a gray-level image and a modified local adaptive thresholding is applied to the contrast-stretched image. In the second step, various morphological operations and Geo-correction method are applied to remove non-text components while retaining the text components. In the third filtering step, non-text components are removed based on the characteristics of each candidate component such as the number of pixels and the bounding box of each connected component Acceptable results have been obtained using the proposed method on 300 domestic news images with a recognition rate of 93.6%. Also, the proposed method gives good performance on the various kinds of images such as foreign news and film videos.
| Original language | English |
|---|---|
| Title of host publication | Document Analysis Systems V - 5th International Workshop, DAS 2002, Proceedings |
| Publisher | Springer Verlag |
| Pages | 341-352 |
| Number of pages | 12 |
| ISBN (Print) | 3540440682, 9783540440680 |
| Publication status | Published - 2002 Jan 1 |
| Event | 5th International Workshop on Document Analysis Systems, DAS 2002 - Princeton, United States Duration: 2002 Aug 19 → 2002 Aug 21 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 2423 |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Other
| Other | 5th International Workshop on Document Analysis Systems, DAS 2002 |
|---|---|
| Country | United States |
| City | Princeton |
| Period | 02/8/19 → 02/8/21 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Theoretical Computer Science
- Computer Science(all)
Cite this
}
Text extraction in digital news video using morphology. / Byun, Hyeran; Jang, Inyoung; Choi, Yeongwoo.
Document Analysis Systems V - 5th International Workshop, DAS 2002, Proceedings. Springer Verlag, 2002. p. 341-352 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2423).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Text extraction in digital news video using morphology
AU - Byun, Hyeran
AU - Jang, Inyoung
AU - Choi, Yeongwoo
PY - 2002/1/1
Y1 - 2002/1/1
N2 - In this paper, a new method is presented to extract both superimposed and embedded scene texts in digital news videos. The algorithm is summarized in the following three steps: preprocessing, extracting candidate regions, and filtering candidate regions. For the first preprocessing step, a color image is converted into a gray-level image and a modified local adaptive thresholding is applied to the contrast-stretched image. In the second step, various morphological operations and Geo-correction method are applied to remove non-text components while retaining the text components. In the third filtering step, non-text components are removed based on the characteristics of each candidate component such as the number of pixels and the bounding box of each connected component Acceptable results have been obtained using the proposed method on 300 domestic news images with a recognition rate of 93.6%. Also, the proposed method gives good performance on the various kinds of images such as foreign news and film videos.
AB - In this paper, a new method is presented to extract both superimposed and embedded scene texts in digital news videos. The algorithm is summarized in the following three steps: preprocessing, extracting candidate regions, and filtering candidate regions. For the first preprocessing step, a color image is converted into a gray-level image and a modified local adaptive thresholding is applied to the contrast-stretched image. In the second step, various morphological operations and Geo-correction method are applied to remove non-text components while retaining the text components. In the third filtering step, non-text components are removed based on the characteristics of each candidate component such as the number of pixels and the bounding box of each connected component Acceptable results have been obtained using the proposed method on 300 domestic news images with a recognition rate of 93.6%. Also, the proposed method gives good performance on the various kinds of images such as foreign news and film videos.
UR - http://www.scopus.com/inward/record.url?scp=33847234482&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=33847234482&partnerID=8YFLogxK
M3 - Conference contribution
SN - 3540440682
SN - 9783540440680
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 341
EP - 352
BT - Document Analysis Systems V - 5th International Workshop, DAS 2002, Proceedings
PB - Springer Verlag
ER -