Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Building a fast digit recognition solution with...

Building a fast digit recognition solution with Python

import all of images in the specific folder, then use the pytesser module to convert the image to string, store it to csv file

Stephen Hsu

May 25, 2013
Tweet

More Decks by Stephen Hsu

Other Decks in Programming

Transcript

  1. Tesseract OCR  Introduction - Open source OCR engine -

    Started at the HP labs between 1985 and 1994 - C, C++ - Google used it for document scan project  Training - Images fonts of Benchmark tools are non-standard. - Training Process is fun !
  2. Technologies  Common  Windows  Linux  Python 2.7

     3rd party APP jTessBoxEditor  Tesseract OCR  Imaging Process  pytesser  PIL  datetime, time  platform  os, sys  csv  glob
  3. Image Pre-Processing  Image Cropping - Crosshairs  Pixel Interpolation

     20M Pixel  99M Pixel  Recognition rate from 42% to 99.99%  Binarization Processing  Lines Recognition
  4. Extracted Text Then ...  Parser In: Number Out: Structured

    number  Store & Convert In: Structured Number Out: CSV File Sample_20130522214241.csv