Core_vision

OMR Auto Grading System using OpenCV that detects, analyzes, and grades MCQ answer sheets in real-time with high accuracy

OMR MCQ Auto Grading System with OpenCV

---

Features

• Real-time webcam scanning
• Image-based processing
• Automatic answer detection
• Perspective correction
• Pixel-based grading
• Score visualization

---

Overview

An intelligent OMR (Optical Mark Recognition) system built with Python and OpenCV that automatically detects and grades MCQ answer sheets in real-time using computer vision techniques.

---

Demo

Input Sheets

Image Processing Pipeline

1️⃣ Grayscale Conversion

Convert image to grayscale for easier processing.

2️⃣ Edge Detection (Canny)

Detect object boundaries in the image.

3️⃣Contour Detection

Identify the answer sheet using the largest contour.

4️⃣threshold

Convert the image to black and white.

5️⃣ Perspective Transformation

The sheet is warped to a top-down view for accurate analysis.

6️⃣ Final Result

requirements Used

• Python
• OpenCV
• NumPy

---

Project Structure

OMR-Auto-Grading-System/
│── main.py
│── utlis.py
│── README.md
│
├── asset/
│   │── original.jpg
│   │── original2.jpg
│   │── original3.jpg
│   │── 2choices_no_answers.jpg
│   │── gray.jpg
│   │── edges.jpg
│   │── contours.jpg
│   │── warped.jpg
│   │── final.jpg
│   │── demo.gif
│
└── Scanned/

---

How to Run

pip install opencv-python numpy
python main.py

---

Controls

• Press 's' → Save result
• Press ESC → Exit

---

Team

• Ganna Amr Emad Eldin — Team Leader
• Habiba Saad Mohamed
• Haneen Mahmoud Abdel Fattah
• Rana Basyouni Askar
• Maryam Teama
• Hana Radwan

---

Academic Supervision

Instructor: * Dr.Marwa Elsiddek

• Eng/Abdelraouf Hawash

Acknowledgment

This project was developed as part of a Computer Vision and Image Processing course.

---

License

For educational purposes only.

---

If you like this project, give it a star!