StudiOCR 0.1.1

Index documents using OCR

StudiOCR

StudiOCR is an application to index notes and make them searchable by using OCR.

• Select .JPG or .PNG files to create a document
• Search through a document to see if it has any matching text
• Any matching text will be highlighted with a colored box based on confidence level

Installation Instructions

Prerequisites

• You must have Qt and Tesseract OCR installed.
• Ubuntu or Debian
  • sudo apt install qt5-default tesseract-ocr
• Arch or Manjaro
  • sudo pacman -S qt5-base tesseract

Install through PyPi

• Optionally, create a fresh venv for installing the package in
  • python3 -m venv venv_StudiOCR
  • source venv_StudiOCR/bin/activate
    • To deactivate the venv run deactivate
• pip install StudiOCR
• Once installed, run StudiOCR to launch the application

Install from Source

• Optionally, create a fresh venv for installing the package dependencies in
  • python3 -m venv venv_StudiOCR
  • source venv_StudiOCR/bin/activate
    • To deactivate the venv, rundeactivate
• git clone https://github.com/BSpwr/StudiOCR
• cd StudiOCR
• pip install -r requirements.txt
• Once installed, run python3 StudiOCR/main.py to launch the application

Usage

Main Window

• Click the Add New Document button to open the add new document window interface
• Click on a document thumbnail (which is generated from the first page) to open the document window interface
• Toggle remove mode to remove existing documents
• Search for a document based on document name by typing in the search bar with the DOC bullet selected
• Search for a document based on matching OCR text by typing in the search bar with the OCR bullet selected

Add New Document Window

• Add/Remove .JPG or .PNG files to be processed by OCR into a document in the database
• Input the document name
• Click on the info icon to display a window explaining document options
• Select the processing model you wish to use: Best (for accuracy) or Fast (for speed)
• Select whether you wish to do image preprocessing (convert to grayscale and increase text contrast)
• PSM Number:

PSM Number	Value
3	Fully automatic page segmentation, but no OSD. (Default)
4	Assume a single column of text of variable sizes.
5	Assume a single uniform block of vertically aligned text.
6	Assume a single uniform block of text.
7	Treat the image as a single text line.
8	Treat the image as a single word.
9	Treat the image as a single word in a circle.
10	Treat the image as a single character.
11	Sparse text. Find as much text as possible in no particular order.
12	Sparse text with OSD.
13	Raw line. Treat the image as a single text line, bypassing hacks that are Tesseract-specific.

Document Window

• Enter text in the search bar to search for matching text in the document
• Click the Next/Previous Page buttons to cycle through the pages in the document
• The current page number is shown at the bottom of the window
• Toggle show matching pages to only display pages with matching text and to cycle through them
• Box Color:

Box Color	Confidence Value
Green	Greater than or equal to 80
Blue	Less than 80 and greater than or equal to 40
Red	Less than 40