Implementation of an Optical Character Recognizer (OCR) for Bengali language

Abstract

Optical character recognition (OCR) is the process of extracting text from an image. The main purpose of an OCR is to make editable documents from existing paper documents or image files. A number of algorithms are required to develop an OCR. Noise removal, skew identification and correction, segmentation, etc are the different steps of developing an OCR. OCR primary works in two phases; they are character and word detection. In case of more sophisticated approach, an OCR also works on sentence detection to preserve documents' structures. In this paper, we would discuss the process of developing an OCR for Bengali language. Lots of efforts have been put on developing an OCR for Bengali. Though some OCRs have been developed, none of them is completely error free. For our thesis, we trained Tesseract OCR engine to develop an OCR for Bengali language. Tesseract is currently the most accurate OCR engine. This engine was developed at HP labs and currently owned by Google. We used a number of software to prepare our training files. Our OCR's library contains 18110 characters and 2617 words. We used "Solaimanlipi" font in our project. We used 200 input files to test the accuracy of our OCR . We are using the latest 3.03 version of Tesseract for windows operating system. For clean image files, the accuracy of our software was as high as 97.56%. It is important to mention that we measured accuracy as the percentage of correct characters and words.