Training an Object Detection Model with TensorFlow OCR Invoice Extractor

How to Run

To use your images instead of the provided ones:

1. Fork and clone this repository to your local machine.

   git@github.com:hrushikesh009/TensorFlow-OCR-Invoice-Extractor.git
   

2. Install required libraries.

   pip3 install -r requirements.txt
   

3. Step 1: Annotate Images

   • Save photos with your objects to ./data/raw.
   • Resize images to (800, 600) with:

     python resize_images.py --raw-dir ./data/raw --save-dir ./data/images --ext jpg --target-size "(800, 600)"
     

   • Train/test split into ./data/images/train and ./data/images/test.
   • Annotate resized images with labelImg and generate xml files.

4. Step 2: Open Colab Notebook

   • Open the Python Notebook (Tensorflow-object-detection-training), which includes a comprehensive guide with a markdown on each stage of building a Tensorflow OCR model.

Enjoy the journey!