DocumentCode :
1634238
Title :
Off-Line Multi-Script Writer Identification Using AR Coefficients
Author :
Garain, Utpal ; Paquet, Thierry
Author_Institution :
Indian Stat. Inst., Kolkata, India
fYear :
2009
Firstpage :
991
Lastpage :
995
Abstract :
The problem of writer identification in a multi-script environment is attempted using a two-dimensional (2D) autoregressive (AR) modeling technique. Each writer is represented by a set of 2D AR model coefficients. A method to estimate AR model coefficients is proposed. This method is applied to an image of text written by a specific writer so that AR coefficients are obtained to characterize the writer. For a given sample, AR coefficients are computed and its L2 distance with each of the stored (writer) prototypes identifies the writer for the sample. The method has been tested on datasets of two different scripts, namely RIMES containing 382 French writers and ISI consisting of samples from 40 Bengali writers. Modeling of writing styles using different context patterns at different image resolution has been investigated. Experimental results show that the technique achieves results comparable with that of the previous approaches.
Keywords :
autoregressive processes; estimation theory; feature extraction; handwriting recognition; image resolution; image sampling; image texture; text analysis; 2D AR model coefficient estimation; Bengali writer; French writer; ISI; L2 distance; RIMES; allograph-based feature extraction; context pattern; handwriting style modeling; image resolution; offline multiscript writer identification; text image sample; texture-based feature extraction; two-dimensional autoregressive modeling technique; Autocorrelation; Context modeling; Feature extraction; Image resolution; Image segmentation; Intersymbol interference; Prototypes; Testing; Text analysis; Writing; 2-D Autoregression; Multi-script; writer identification;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Document Analysis and Recognition, 2009. ICDAR '09. 10th International Conference on
Conference_Location :
Barcelona
ISSN :
1520-5363
Print_ISBN :
978-1-4244-4500-4
Electronic_ISBN :
1520-5363
Type :
conf
DOI :
10.1109/ICDAR.2009.222
Filename :
5277544
Link To Document :
بازگشت