Abstract
Document analysis tasks such as pattern recognition, word spotting or segmentation, require comprehensive databases for training and validation. Not only variations in writing style but also the used list of words is of importance in the case that training samples should reflect the input of a specific area of application. However, generation of training samples is expensive in the sense of manpower and time, particularly if complete text pages including complex ground truth are required. This is why there is a lack of such databases, especially for Arabic, the second most popular language. However, Arabic handwriting recognition involves different preprocessing, segmentation and recognition methods. Each requires particular ground truth or samples to enable optimal training and validation, which are often not covered by the currently available databases. To overcome this issue, we propose a system that synthesizes Arabic handwritten words and text pages and generates corresponding detailed ground truth. We use these syntheses to validate a new, segmentation based system that recognizes handwritten Arabic words. We found that a modification of an Active Shape Model based character classifiers—that we proposed earlier—improves the word recognition accuracy. Further improvements are achieved, by using a vocabulary of the 50,000 most common Arabic words for error correction.
Highlights
Modern document analysis heavily depends on automated processes as pattern recognition or segmentation
Using Support Vector Machines (SVMs) we achieved good results (95.4% ± 0.3% and 97.14% ± 0.06%), which are similar to our Active Shape Models (ASMs) based approach
We have presented an efficient approach to synthesize Arabic handwritten words and text pages from Unicode
Summary
Modern document analysis heavily depends on automated processes as pattern recognition or segmentation These processes need to be trained using a database and validated using corresponding, suitable ground truth (GT), though. The IESK-arDB database, that we proposed in [6], contains international town names and common terms including GT for segmentation. Since both databases are limited by the number of samples and words and contain single words or small sentences only, we believe automatized generation of databases customized for specific research is a helpful complement. Advantages are that samples can be created and quickly for any word or text at any time, detailed GT are created simultaneously
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have