Optical Character Recognition, or OCR, is a technology that can copy machine-printed or handwritten text from an image or a PDF file and output it on a computer like typed text, that can be edited or searched using an engine. The output can be in any of the DOC, RTF, XLS, PDF, HTML or TXT formats.
OCR tools at the home or small scale level is usually available in two types. They can be installed as applications on a PC or as online services on websites. Many computer users might already have an OCR tool on their PCs, as it is available in the Microsoft OneNote.
The online tools require the user to upload files onto a server and the converted file is transferred back. This method usually offers better accuracy of conversion but have a file size limit and the number of pages that can be converted for free. Obviously, it is not advisable to convert sensitive or confidential files in this manner; we don’t know if the server stores or peep into your file.
Desktop applications, of course, do not have this risk and will convert large files without any hassle. However, the accuracy of conversion depends on the software. Most desktop and online OCR tools can convert characters from 20 to 40 languages.
Perhaps the most widely used free OCR tool online is the well known Google Docs which uses the OCR technology with Google to computerise books. The file size limit is 2 MB and the conversion is quite fair.
Another online OCR tool is at OnlineOCR.net, which has better conversion accuracy and has file size limit of 5 MB which can be converted into DOC, XLS or TXT formats. It allows converting 15 images per hour.
Similarly, the Free OCR on Free-Ocr.com can process a file of up to 6 MB. The Free Online OCR tool at NewOcr.com can convert large files running into several pages.
Among the desktop OCR tools, the most popular are probably SimpleOCR by SimpleSoftware and Free OCR to Word available at OcrToWord.com.
It has about 98% accuracy. Most OCR tools are based on the Tesseract engine. The FreeOCR software available at PapaerFile.net may require Microsoft .NET to function properly.
For Mac users a PDF OCR X is available by Web Lite Solutions. It is a simple imple drag-and-drop utility but can convert only single page PDF files. Premium purchase is required for multi-page conversions.
If you are using a scanner to scan images, you can try setting it to 300 dpi, colour JPEG or 300 dpi, gray-scale JPG. These settings should give good results. Getting decent accuracy for files scanned using a digital camera is difficult. It is better to use a smartphone camera and take snapshot using applications like CamScanner.
Leave a Reply