DefinePK hosts the largest index of Pakistani journals, research articles, news headlines, and videos. It also offers chapter-level book search.
Title: An advanced, robust binarization and character recognition in document images based on raspberry Pi
Authors: Anitta Vincent, Gincy Varghese
Journal: ARPN Journal of Engineering and Applied Sciences
Publisher: Khyber Medical College, Peshawar
Country: Pakistan
Year: 2015
Volume: 10
Issue: 17
Language: English
Digitization is a major scenario through which transactions and information sharing occur very fast. The notion of ‘paperless office’ is gradually coming into existence. Besides, there are a plethora of treasured documents which need to be safeguarded. In such a plot, an adequate system for reclamation and digital conservation of documents becomes indispensable. Here, a robust system for binarization and character recognition of deteriorated document images is proposed. Segmentation of text from deteriorated document images is a very tough task due to the high variation within and between the document background and the foreground text in different document images. The binarization technique employed here is a novel technique which utilizes a modified adaptive image contrast based method. This method can tolerate non-uniform background and various artifacts that creep in. Next step is the image segmentation, followed by character recognition using Artificial Neural Network. The entire system is implemented on a low cost minicomputer board named Raspberry Pi.
Loading PDF...
Loading Statistics...