Auto Evaluation System for True/False Answer Sheets NCVPRIPG24

Team Name - AI Avengers

• One of the top-5 teams in the NCVPRIPG'24 Summer Challenge.

Members -

• Akshat Jain (B22CS007)
  
• Gaurav Manish (B22CS079)
  
• Mukund Gupta (B22CS086)
  
• Rutam Rajhansa (B22ME055)
  

Overview

This project involves an automated evaluation system designed to grade true/false answer sheets from images. The system ensures the correct orientation of the answer sheets, filters out unwanted text, and extracts the answers using OCR (Optical Character Recognition) technologies.

Features

• Automated Image Orientation Correction: Utilizes PyTesseract to correct the orientation of the input images.
• Image Processing: Applies Gaussian blur to focus on the answer areas and filter out unwanted text.
• Optical Character Recognition: Uses EasyOCR to accurately read and extract answers from the processed images.
• Validation and Error Handling: Implements comprehensive checks to ensure the accuracy and reliability of the system.

Technologies Used

• Python: Primary programming language.
• PyTesseract: For OCR-based image orientation correction.
• OpenCV: For image processing tasks, including applying Gaussian blur.
• EasyOCR: For extracting text from images.
• NumPy: For numerical operations and array handling.
• Matplotlib: For visualizing and debugging image processing steps.

Installation

1. Clone the Repository

   git clone https://github.com/yourusername/auto-evaluation-system.git
   cd auto-evaluation-system