HighlightedTextDetector

Extracts highlighted text from the image of a page

Process an image of a page taken by a handheld device to detect the highlighted portion of the text, and then running OCR for extracting the highlighted text.

Running The Project

1. Clone the project from GitHub and import (File > Open) it into PyCharm.

2. Now install (File > Settings > Project > Python Interpreter > [+] button) following modules into the project: opencv-python, google-cloud-vision, numpy, and Pillow. You may also use commands:

pip install opencv-python 
pip install google-cloud-vision
pip install numpy
pip install Pillow

3. Setup Google Vision Server:

   a. Create Project: Login into your Google account > Open the Google-Cloud page > Create a project with name- 'HighlightedTextDetector' (you may give any name).

   b. Enable Vision API for The Project & Download Credentials: Go to Vision-Api page > Select your project 'HighlightedTextDetector' > Continue > Are you planning to use this API with App Engine: No > What Credentials do I need? > Service account name: HighlightedTextDetector > Role: Owner > Key type: JSON > Continue > Save (Now, place this JSON file somewhere safe; we will need it for authentication)

   c. Billing Account (Free): Open Billing Page > Add Billing Account > Enter billing account details > Continue > Enter credit-card info > Submit and Enable Billing

For details, visit the official guide.

4. Open main.py > update the variable GOOGLE_VISION_API_KEY_PATH with your API Key (downloaded JSON) file path > then run processStepByStep.py