Visual Guard

Visual Guard is a facial recognition system designed to securely identify students by comparing their faces against pre-stored data. Ideal for educational institutions and government agencies, Visual Guard enhances student authentication processes, offering real-time recognition and comprehensive student data verification.

Features

• Facial Recognition: Captures student faces via live camera or uploaded images and matches them with pre-registered data.

• Real-Time Processing: Utilizes Flask-SocketIO for instant feedback and processing.

• Student Data Verification: Matches faces with data fields like Name, Age, Registration Number, etc., stored in users.json.

• Flexible Image Input: Accepts Base64-encoded images from cameras or file uploads.

• Optimized for Efficiency: Fast image resizing and preprocessing for large images.

• Error Handling: Returns specific error messages for failed image decoding, face detection, or recognition.

Technology Stack

• Flask: Web framework for server-side operations.

• Flask-SocketIO: Enables real-time server-client communication.

• face_recognition (dlib): Library for face detection and recognition.

• OpenCV (cv2): For image processing.

• NumPy: Image data handling.

• Pillow (PIL): Image manipulation.

System Workflow

1. User Data Initialization: Loads student data and facial encodings from users.json.

2. Real-Time Image Processing: Processes submitted images, resizing them and extracting face encodings.

3. Identity Verification: Compares face encodings with stored data to confirm identity.

4. Response to Client: Returns the student profile or an error message in real-time.

Installation and Setup

Prerequisites

• Python 3.12

• pip for installing dependencies

Installation Steps

1. bashCopy code

  git clone cd visual-guard

3. bashCopy code

pip install -r requirements.txt

3. Prepare User Data: Add student details in users.json. Include paths to student images.

4. Run the Application:

python app.py

Sample users.json Structure

json

{
  "Name": "L",
  "Age": 19,
  "Designation": "Student",
  "Registration No.": "PT22CS092"
}

Error Handling

• Invalid Image Format: "Failed to decode image."

• No Face Detected: "No face found in the image."

• Recognition Failure: "Could not encode face features."

• No Match: "No match found in database."

Future Enhancements

• Centralized student databases

• Enhanced AI-based recognition

• Mobile and cloud integration

Documentation

For detailed project documentation, click here.