So the OCR works
Imagine that you have received from a colleague a presentation in a paper format. You don't want to edit them on the PC, since individual passages like that. Consequently, you have to scan it and call your OCR program. Now the following happens:
- The Software first makes a so-called layout analysis. Ensure you look at the design of the page and separates images from Text. She also remembers the Position on the page. Then the number of paragraphs to be counted, and individual elements such as page numbers are stored.
- Now comes the hard part. The Software looks at the individual blocks of text, and disassembled these in turn into sentences. The sentences are then split into individual words and words into letters.
- The OCR Software includes a pattern of letters and characters. The program compares the scanned letters now with these Patterns. If you are similar to 99% is decided by the algorithm, that it must be likely to this letter. Here, he is very precise, because he can compare many samples in a short period of time. So the distinction between "8" and "B succeeds in successfully".
- The letters and characters are recognized. Then they are combined as words, and back in its place in the set of classified. Once the Software is finished, you will be saved All in a normal document, which you can then edit. Done!
OCR behind it
Do you want to test your Software Knowledge? Make you with our great Software-Quiz!
