DocumentCode :
3475389
Title :
Gradient-based video text localization algorithm with statistical analysis of text-like features
Author :
Cheng-Hsuan Li ; Yung-Lin Huang ; Shao-Yi Chien
Author_Institution :
Dept. of Electr. Eng., Nat. Taiwan Univ., Taipei, Taiwan
fYear :
2011
fDate :
6-8 Sept. 2011
Firstpage :
340
Lastpage :
341
Abstract :
Video processing techniques in most of the current display systems are performed without considering the text information. The proposed algorithm aims to localize the text regions for video processing. To derive the robust text-like features, a statistical analysis is introduced. With the 2D structure analysis and multi-resolution refinement, most text regions are localized using the proposed text-like feature.
Keywords :
character recognition; feature extraction; image resolution; statistical analysis; video signal processing; 2D structure analysis; gradient-based video text localization algorithm; multiresolution refinement; statistical analysis; text information; text regions; text-like features; video processing; Algorithm design and analysis; Feature extraction; Generators; Laplace equations; Multimedia communication; Statistical analysis; Streaming media;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Consumer Electronics - Berlin (ICCE-Berlin), 2011 IEEE International Conference on
Conference_Location :
Berlin
Print_ISBN :
978-1-4577-0233-4
Type :
conf
DOI :
10.1109/ICCE-Berlin.2011.6031864
Filename :
6031864
Link To Document :
بازگشت