Project Description:

The IoT Pet Feeder with Face Recognition is an intelligent, internet-connected system that ensures only a specific pet is fed at the right time and in the right quantity. This project combines IoT hardware, face recognition AI/ML models, and a web interface for monitoring and control. It is ideal for households with multiple pets, ensuring that food isn’t stolen by the wrong pet or wasted.

Key Components:

Face Recognition System:

• A camera connected to a Raspberry Pi or NodeMCU detects the pet approaching.

• Using an ML model (e.g., OpenCV + TensorFlow or face-api.js), it identifies the pet.

• Only if the recognized pet matches the stored profile, the feeder is activated.

Web Dashboard (Frontend):

• Built using HTML, CSS, Bootstrap, JavaScript.

• Displays:

  • Live pet camera feed.

  • Feeding logs (time, pet recognized).

  • Manual override to feed.

  • Profile settings for pets (image, schedule, quantity).

Backend (API and Logic):

• Built using Node.js, PHP, or Java.

• Handles:

  • Image storage and pet profiles.

  • Communication with the IoT device (via MQTT/HTTP).

  • Feeding schedule logic and authentication.

  • Feeding event logs.

Mobile/Web Notifications:

• Sends alerts when:

  • A pet has been fed.

  • A new face is detected.

  • Food container is low (sensor-based).

ML Model Details:

• Train a lightweight face recognition model using labeled images of each pet.

• Use KNN, CNN, or Haar Cascades for image classification.

• Optimized to run on edge devices or offload processing to the server.

IoT Hardware Required:

• Raspberry Pi or NodeMCU

• Servo motor (for controlling food release)

• Camera module

• Load sensor (to detect food level)

• Wi-Fi module

• Power source

Security Features:

• Secure login for owners/admins.

• HTTPS communication between dashboard and backend.

• Optional facial data encryption for privacy.