Optical Character Recognition – OCR
Optical Character Recognition or OCR is the process of reading or detecting texts from images, pdf files, scanned images, text files, etc. This technology is a huge leap in the field of optical science and automation. Rather than entering textual data manually, OCR is being used nowadays for quicker and efficient output. More often than not, we encounter sitwwwions where our photo identity and address proofs, etc are asked for. They are then scanned and uploaded into the system, from where the data gets automatically transferred into the database.
The usage of OCR is manifold. Data read with OCR can be searched, edited, stored or used for some other purpose as well. For example, the Smartphone’s now has the capability of converting text to speech with the help of OCR only. Translating of scripts, e-books and articles using this technology is rapidly becoming familiar and its significance lies in the fact that almost all the recent technological creations has this feature.
docEdge DMS provide you an end-to-end document management solution, which includes scanning your documents. We also provide professional scanning services that usually produces images in TIFF format. The main advantage of the TIFF image format is that it retains image quality – TIFF uses “lossless” compression. This makes it easier for Optical Character Recognition (OCR) software to “read” the text in the scanned images more accurately. With help of Google Vision API is one of the first DMS solutions to offer Hindi OCR.
OCR Module usually follows a 3 step process:
- Text is extracted from TIFF images and a plain text file is created (often in a variant of HTML so that formatting and placement is retained).
- The TIFF file is converted into a PDF file, which usually takes up less disk space.
- The text file created in step 1 is embedded in the PDF file, with the words placed at the correct spot. The text now becomes selectable in the PDF and can read by machines.
docEdge DMS is able to read the text from the PDF files generated in step 3 above and the search engine indexes the text. This makes full text keyword search possible against OCRed images.
docEdge DMS comes with built-in OCR support as well as integration with Google Vision API – AI/ML based OCR platform. Here, Google Vision API support almost all kind of languages. Rules can be setup for running OCR in bulk across a large number of uploaded files, or individual files can be uploaded with OCR being run in real time.