Skip to content

Latest commit

 

History

History
37 lines (21 loc) · 1.29 KB

README.md

File metadata and controls

37 lines (21 loc) · 1.29 KB

Form Character Recognizer

Web App screenshot!

Simple Script for Extracting and Recognizing handwriten or written numbers from given form.

Instructions

Quick Start

Prerequisite

  • python 3+
  • opencv
  • numpy

Usage

Make a directory called Forms where the script lives on your computer, copy forms into that, and run the script like so:

python3 generate_dataset.py

Details

main image!

After Reading a form image, as forms may be rotated duo to scanning process, images get horizontal by detecting and calculating images rotated angle using warpAffine function. after that, black boxes (eyes) on the sides of the page representing a row are detected using findContours function. to ensure that found boxes are the one we want, after appling a blur filter, closing operation, and a threshold, their area are examined to match a known range.

After detecting rows by sorting found contours, cause the distance between cells are certain, handwritten numbers' cell and numbers written in letters' cell and just the same, box of each number is found.