Basic knowledge of OCR

ocr,optical character recognition abbreviation, that is, optical recognition system, is a branch of graphics recognition, OCR is for the printing character, the use of optical way to convert the document data into the original data black and white lattice image file, The recognition software identifies the text in the image as Chinese and English, and then converts it into a text format. So the computer can be recognized by OCR system, you can clearly see what you saw, read what, especially the text material.

(1) Processing process:
Image input, image pre-processing, text feature extraction, and contrast recognition finally, by artificial correction, the text corrections will be false, and finally output the result.

(2) Specific process

1, image input: The target file after optical instrument generated bitmap stored in the computer.

2, image preprocessing: including two value, corrosion and expansion, median filter.

Image binary: Only two colors (often black and white) are visible when the image is displayed.

3, character extraction: According to the characteristics of characters, in different regions to collect data.

4. Compare database: Make the corresponding letter and number template by the artboard as the standard of comparison.

5, comparative recognition: Based on the characteristics of the use of the comparison, find the most similar characters.

6, output: the most similar character as the result of recognition output, that is, the identification code output.