Digital Prank

Optical Character Recognition popularly known as OCR, is a technology that helps in translating images carrying text (printed text Images) in to text characters. OCR uses pattern recognition to extract the text out of the images and this technology can help you to scan your text book to a editable document in your computer with ease. OCR has now part of commercial business extending its hands in the line of data entry and bill processing giving higher work productivity.  There are lot of software available in the market to do the job but when it comes to programming in C, C++ or especially in java one can find limited libraries to achieve the same programmatically.

Opensource libraries for OCR are very few for commercial use and for java there are literally none as of today except for few commercial ones like Asprise OCR SDK library. Asprise OCR Java SDK library trial version can be downloaded freely from Internet and its looks good and solid with respect to accuracy. One of my friend was taking to me about a open source library called JavaOCR which has gone missing now.

There are few non java open source libraries like GOCR and tesseract-ocr exists which are quite good. GOCR is pretty famous and is primarily for C++ but can be interfaced with Java by converting the images to PPM using nConvert but it will not help you for a big application with vast functionalities. There is a related thread in forums.sun.com for your reference. Google’s Tesseract is another API mainly for C/C++ and it is not for Java unless you do something with java runtime. So I strongly believe it is better to go for a commercial library to achieve OCR rather than for a open source complex solution. In this case open source solution makes the maintainability tougher due to lack of technical support on the open source side.