Title :
Blind Separation of Mixed and Disturbed Correlated Ancient Texts
Author :
Posht-Panah, Shahrouz ; Sahaf, Masoud Reza Aghabozorgi
Author_Institution :
Electr. & Comput. Eng. Dept., Yazd Univ., Yazd, Iran
Abstract :
One of problems that has been investigated and concentrated in image processing area is separation and recognition of ancient documents and texts images that have been mixed and disturbed in consequence of passing many times. In recent years ICA has been used for solving this problem, but essential assumption in ICA is independence of sources, whereas in some problems sources are not independent. So in this paper it has been tried, by utilizing ESPRIT and MUSIC algorithms from blind source separation techniques, to propose a method for separation and recognition of correlated ancient documents and texts that have been mixed. The good performance of this method has investigated for real images.
Keywords :
blind source separation; image classification; independent component analysis; text analysis; ESPRIT algorithms; MUSIC algorithms; blind source separation techniques; disturbed correlated ancient texts; image processing; independent component analysis; mixed correlated ancient texts; texts images; Algorithm design and analysis; Blind source separation; Image processing; Image recognition; Independent component analysis; Ink; Multiple signal classification; Signal processing algorithms; Source separation; Text recognition; Blind Source Separation(BSS); Independent Component Analysis(ICA); MUSIC and ESPRIT algorithms; Recognition of ancient documents;
Conference_Titel :
Computer and Electrical Engineering, 2009. ICCEE '09. Second International Conference on
Conference_Location :
Dubai
Print_ISBN :
978-1-4244-5365-8
Electronic_ISBN :
978-0-7695-3925-6
DOI :
10.1109/ICCEE.2009.120
