DocumentCode
2638835
Title
Instantaneously responsive subtitle localization and classification for TV applications
Author
Zafarifar, Bahman ; Cao, Jingyue ; De With, Peter H N
fYear
2011
fDate
9-12 Jan. 2011
Firstpage
165
Lastpage
166
Abstract
This paper presents an algorithm for localization and classification of subtitles in TV videos. We extend an existing static-region detector with object-based adaptive filtering and binary classification of subtitle bounding boxes, using geometry and text-stroke alignment features. Compared to this static-region detector, we reduce the number of falsely detected subtitle pixels by a factor of 20, at the cost of only 2% of subtitle pixels.
Keywords
adaptive filters; feature extraction; image classification; video signal processing; TV applications; TV videos; binary classification; geometry; object-based adaptive filtering; pixel detection; static-region detector; subtitle bounding boxes; subtitle classification; subtitle localization; text-stroke alignment features; Classification algorithms; Detectors; Filtering; Geometry; Pixel; TV; Videos;
fLanguage
English
Publisher
ieee
Conference_Titel
Consumer Electronics (ICCE), 2011 IEEE International Conference on
Conference_Location
Las Vegas, NV
ISSN
2158-3994
Print_ISBN
978-1-4244-8711-0
Type
conf
DOI
10.1109/ICCE.2011.5722519
Filename
5722519
Link To Document