OCR (Optical Character Recognition) explained.

Posted on : 2017-07-09 in technology
By Ajithkumar

Well, everyone feels lazy to type or copy the things which are already in a paper format so in order to solve this problem, researchers and scientists came with a solution called as OCR.
So today lets see what is OCR (Optical Character Recognition).

Basically, OCR is similar to the scanner the only difference is scanner converts your hard copy into an image. But in the case of OCR, it converts your hard copy into an editable word document.

OCR is available both as a machine and in recent days there are many apps and software that can convert your image into text. One major example is Google Drive, which offers this method to convert your any type of file into a text document and it uses AI for this process.
Now let's see how does it work.

When we keep a piece of document or handwritten notes into the machine or upload it in case of software. First, it removes all the background colors and convert it completely into black and white so it will be easy for it to recognize the text. Next using its font library it matches the closest font to that in the paper, then converts it totally into that specific font.

Now you may wonder how it recognizes each and every letter as everyone's handwriting differs at least in a small way. Here the AI comes into the play it scans each and every pixel of the letter and recognizes all the curves and strokes of each letter, then compares it with the letters of the font then it decides.

Well, even this process is not 100% accurate these days. And even the document should have a legit handwriting so you can get better results. In order to improve the output nowadays OCR uses its dictionary to see whether that particular word is relevant to that sentence. In this way, we get the output as a text document and then we can edit the document.
Hope you people like the post if so please do share it.


Other posts you may like