PDF To Text

Introduction

System converts scanned documents in PDF format to text. The steps involve converting the PDF pages to images and performing OCR on it.

Development Environment

• Python 3.5.3
• OpenCV 3.4.0
• ImageMagick-6.9.9-43-Q8-x64
• Python wand
• Tesseract 4
• pytesseract 0.2.2
• Ghostscript 9.23 Above are the list of key python packages used for development.

Reference links

http://docs.wand-py.org/en/latest/guide/install.html#install-imagemagick-on-window https://www.imagemagick.org/download/binaries/ [ImageMagick-6.9.9-43-Q8-x64-dll.exe] https://github.com/tesseract-ocr/tesseract/wiki/4.0-with-LSTM#400-alpha-for-windows http://digi.bib.uni-mannheim.de/tesseract/tesseract-ocr-setup-4.00.00dev.exe https://ghostscript.com/download/gsdnld.html

Usage and Deployment

Navigate to the root directory PdfToText and use following command to invoke the program. PTTConfig.properties property file controls program execution

python src\Pipeline.py config\PTTConfig.properties

Configuration parameters

Mandatory Parameters

• logFile = log/ENLog.log, Path of the log file
• inputPath = input/ Path where input pdf are stored
• tempPath = temp/ Path to store the temporary files
• outputPath = output/ Path where output will be stored
• magickHome = C:/Program Files/ImageMagick-6.9.9-Q8/ Absolute path of Image Magic installation directory
• tesseractHome = C:/Program Files (x86)/Tesseract-OCR/ Absolute path of Tesseract installation directory
• tessData = C:/Program Files (x86)/Tesseract-OCR/tessdata/ Absolute path of directory where Tesseract language models are stored Please give the trailing slash / while specifying the path
• resolution = 400 , Integer value representing dpi with which images need to be generated, Higher the value better the image quality, But increasing it beyond a point doesn't help.
• outputFormat = tsv , Possible values are pdf, hocr, txt, tsv

Optional parameters

• logLevel = INFO, Level of logging required. Possible values are CRITICAL, ERROR, WARNING, INFO, DEBUG, NOTSET. https://docs.python.org/3.6/library/logging.html#levels

Directory Structure

• PdfToText [Root Directory]
  • config [All .property files stays here]
  • input [Input PDF files]
  • log [Default log location]
  • output [Output OCR file stays here]
  • src [Python source folder]
  • temp [Temporary storage location]

This document was edited at https://dillinger.io/