Optical Mark Recognition (OMR) MCQ Automated Grading

Technologies Used

Work Pipeline

For Educational Purposes ONLY

The code works as follows:

• Reads the original image in grayscale.
• Finds the edges.
• Finds the contours, from which we extract the biggest rectangle's cornerpoints.
• The image is processed using warp to get the bird's eye view.
• Perform a threshold
• Find the shaded/marked boxes/bubbles
• Generates an answer key from a solution image
  • This is an option step, as we can define any answer key in a specific format.
• Saves the grades in a .csv file

---

Install

Clone the repository, or alternatively download it

git clone https://github.com/Bakhomious/OMR-MCQ-Automated-Grading
cd OMR-MCQ-Automated-Grading

Install pip requirements

python -m pip install --user -r requirements.txt

Run

python main.py