This Python project utilizes image processing and machine learning techniques to automate the extraction and prediction of tabular data from images. The implementation achieves over 80% accuracy and is designed to be versatile for integration into various projects.
The Jupyter Notebook in this repository demonstrates the step-by-step process of image preprocessing, contour extraction, digit recognition, and prediction. It involves techniques such as Gaussian blur, adaptive thresholding, contour detection, and Tesseract OCR for recognizing Arabic digits. The project also incorporates a machine learning model to predict the values of the extracted digits.
Make sure you have the following dependencies installed:
- Python 3.9.12
- OpenCV
- NumPy
- Matplotlib
- Tesseract OCR
- PyTesseract
- PIL (Pillow)
- Scikit-learn
You can install these dependencies using:
pip install -r requirements.txt
- Open the Jupyter Notebook Automated_Tabular_Extraction_and_Prediction.ipynb.
- Follow the step-by-step guide to preprocess images, extract contours, recognize digits, and predict values.
- Customize the notebook for your specific use case or integrate the provided functions into your projects.
The project demonstrates accurate tabular data extraction and prediction, with a focus on Arabic digits. Evaluate the results and tailor the model parameters or preprocessing steps based on your requirements.
This project leverages the power of OpenCV, Tesseract OCR, and machine learning. Special thanks to the open-source community for these invaluable contributions.
Feel free to explore, modify, and integrate this solution into your projects. Happy coding!