InnoVision

🛠️ AI-Powered Image Analysis of Industrial Objects for Quality Assurance

An end-to-end automated quality inspection system using YOLOv8, Python, and Raspberry Pi, capable of identifying defective industrial components in real-time and rejecting them automatically with a servo-controlled mechanism.

---

📸 Demo

(Defected Product (Gear) detection and automatic ejection demonstration)

---

🚀 Features

• 🔍 Real-time object detection and defect classification using YOLOv8
• 📷 Live video feed processing from a USB camera
• 🤖 Automated rejection system using a servo motor
• ⚙️ Conveyor control using 12V DC motor and ultrasonic sensor
• 🧠 Edge device deployment on Raspberry Pi 4
• 💡 Cost-effective solution for industrial automation

---

🧠 How It Works

1. Object Detection
   A USB camera captures a live feed as objects move on a conveyor belt.

2. YOLOv8 Analysis
   Each frame is processed using YOLOv8 to classify the object as GOOD or DEFECTIVE.

3. Automation Logic

   • If GOOD: The object continues on the conveyor.
   • If DEFECTIVE: A servo motor pushes it off the belt.

4. Ultrasonic Sensor
   Detects the presence of an object to trigger image capture and model inference.

---

🔧 Hardware Used

Component	Description
Raspberry Pi 4	Main controller
USB Camera	Live feed acquisition
12V DC Motor	Conveyor drive
Ultrasonic Sensor	Object detection trigger
Servo Motor	Defective object rejection
L298N Motor Driver	Controls DC and servo motors
Conveyor Belt	Transports objects

---

💻 Software Stack

• Python 3.9+
• YOLOv8 (Ultralytics)
• OpenCV
• GPIO Zero / RPi.GPIO
• Linux (Raspberry Pi OS)

---