🎯 Bingo Detection – Image Processing Model

Automated Bingo Card Recognition & Validation in MATLAB

---

📌 Project Overview

This project is an image processing–based Bingo card detection and validation system, developed as part of an academic assignment. The system analyzes an input image of a Bingo card, detects stamped cells, corrects rotation if necessary, digitizes the card into a logical grid, and determines whether a valid Bingo pattern exists.

The project demonstrates practical use of MATLAB image processing techniques, including frequency-domain analysis, geometric correction, shape detection, and logical validation. The solution is designed in a modular and readable way, making the processing pipeline transparent and easy to understand.

---

🧠 Problem Description

Scanned or photographed Bingo cards often suffer from:

• Rotation or skew
• Uneven lighting
• Background noise
• Imperfect stamp placement

The objective of this assignment is to build a custom optical recognition system that can:

• Automatically straighten a rotated Bingo card
• Detect circular stamps placed by the player
• Map visual detections to a 5×5 digital grid
• Check for a Bingo (row, column, or diagonal)
• Display intermediate processing steps visually

The output includes both terminal output (digitized array + Bingo status) and graphical visualization.

---

🖼️ Visual Processing Stages

The assignment requires the processing function to present a figure composed of multiple subplots, each illustrating a key step of the pipeline.

---

Subplot Breakdown

1. Original Image
   Raw input image of the Bingo card.

2. Fourier Magnitude of Edge Image
   Shows frequency-domain representation used to estimate the card’s rotation angle.

3. Straightened and Cropped Image
   The image after rotation correction and cropping to the Bingo grid.

4. Hough Space – Line Detection
   Visualization of detected grid lines used to infer the Bingo grid structure.

5. Hough Space – Circle Detection
   Visualization of detected circular stamps.

6. Final Overlay Visualization
   Grayscale image with detected circles marked. If a Bingo exists, the winning row/column/diagonal is drawn with a green line.

---

⚙️ System Workflow

The system processes each image using the following steps:

1. Preprocessing & Masking
   Filters out unnecessary image regions and enhances relevant features.

2. Rotation Estimation (FFT-based)
   Uses Fourier transform magnitude to detect dominant orientation.

3. Rotation Correction & Cropping
   Aligns the Bingo card to a canonical orientation.

4. Grid Line Detection
   Applies a Hough-like transformation to detect grid lines.

5. Circle Detection
   Detects circular Bingo stamps using geometric and Hough-based methods.

6. Digitization
   Maps detected circles to a 5×5 logical matrix.

7. Bingo Validation
   Checks rows, columns, and diagonals for a valid Bingo pattern.

---

▶️ How to Run the Project

Main Entry Point

The main script of the project is:

main.m

Steps

1. Open MATLAB
2. Set the project root directory as the current folder
3. Run:

main

The script automatically loops through all images in the input folder and processes them sequentially.

---

📂 File Structure

├── main.m                 % Main controller script
├── scanner.m              % Core processing function
├── findRotationAngle.m    % Rotation estimation using FFT
├── findCircle.m           % Circular stamp detection
├── findBingoLine.m        % Identifies winning Bingo line
├── checkBingo.m           % Validates Bingo rules
├── README.md              % Project documentation
└── input/                 % Folder containing Bingo card images

---

🧩 File Responsibilities

main.m

• Iterates through the input image folder
• Calls the processing pipeline
• Displays results and prints output

scanner.m

• Core required function
• Input: image matrix
• Output: 5×5 digitized matrix + Bingo boolean
• Produces visualization subplots

findRotationAngle.m

• Estimates image rotation using frequency-domain analysis

findCircle.m

• Detects circular Bingo stamps

findBingoLine.m

• Determines which row/column/diagonal forms a Bingo

checkBingo.m

• Applies logical Bingo rules and returns final decision

---

🧪 Output

For each processed image, the system:

• Prints the digitized 5×5 matrix to the terminal
• Prints whether a Bingo exists
• Displays a multi-panel visualization of processing steps

---

🎓 Evaluation

This project received full marks (100/100) for successfully fulfilling all functional, visual, and structural requirements of the assignment, including correct digitization, rotation handling, Bingo detection, and clear visual presentation.

---

Author: Shihab Muhtasim

Course: Basic Image Processing Algorithms, IPCVAI, PPCU, Budapest, Hungary

Language: MATLAB