👁️ Computer Vision Projects Sandbox

A collection of Computer Vision (CV) projects exploring Human-Computer Interaction (HCI), real-time object detection, and AI-powered automation using Python, OpenCV, MediaPipe, and YOLO.

This repository currently contains:

• 🎵 Hand Gesture Music Controller
• 🚗 YOLO Real-Time Vehicle Counter

---

📂 Repository Structure

.
├── hand_music_controller.py
├── car_counter.py
├── requirements.txt
├── mysong.mp3
├── yolov8n.pt
└── README.md

---

🚀 Features

🎵 Hand Gesture Music Controller

• Real-time hand tracking using MediaPipe
• Detects finger positions from webcam input
• Plays music when an open hand is detected
• Stops music when a closed fist is detected
• Lightweight and CPU-friendly

🚗 YOLO Vehicle Counter

• Real-time vehicle detection using YOLOv8 Nano

• Detects:

  • 🚗 Cars
  • 🏍️ Motorcycles
  • 🚌 Buses
  • 🚚 Trucks

• Live bounding boxes

• Vehicle counting

• Works with webcam or video files

• Optimized for CPU execution

---

🛠️ Environment Setup

Prerequisites

Install the required system packages:

sudo apt update

sudo apt install \
python3-venv \
python3-pip \
libgl1-mesa-glx \
libasound2-dev -y

---

Create a Virtual Environment

python3 -m venv cv_env

Activate it:

Linux/macOS

source cv_env/bin/activate

Windows

cv_env\Scripts\activate

---

Install Dependencies

pip install -r requirements.txt

---

🎵 Project 1 — Hand Gesture Music Controller

Control music playback using simple hand gestures captured by your webcam.

Supported Gestures

Gesture	Action
✋ Open Hand	Play music
✊ Closed Fist	Stop music

---

How It Works

MediaPipe detects 21 hand landmarks.

Finger states are determined using landmark coordinates.

Because image coordinates start at the top-left corner, a raised finger has a smaller Y coordinate than its lower joint.

The rule used is:

Finger Tip Y < Finger PIP Y
        ↓
Finger is UP

---

Run the Application

Place your music file in the project directory:

mysong.mp3

Then run:

python hand_music_controller.py

---

🚗 Project 2 — YOLO Real-Time Vehicle Counter

A lightweight vehicle detection system powered by YOLOv8 Nano.

---

Detection Pipeline

1. Capture webcam/video frame
2. Run YOLOv8 inference
3. Filter vehicle classes
4. Draw bounding boxes
5. Display live vehicle count

Supported COCO classes:

Class	ID
Car	2
Motorcycle	3
Bus	5
Truck	7

---

Run the Vehicle Counter

python car_counter.py

To use a video instead of the webcam, change:

source = 0

to:

source = "traffic.mp4"

---

📦 Technologies Used

• Python
• OpenCV
• MediaPipe
• YOLOv8 (Ultralytics)
• NumPy
• Pygame

---

⚡ CPU Optimization

This repository is designed to work efficiently on systems without NVIDIA GPUs.

Benefits include:

• Smaller installation size
• Reduced dependency footprint
• Faster setup
• Lower memory usage
• Ideal for laptops and virtual machines

---

🔮 Future Improvements

Hand Gesture Controller

• Swipe gesture detection
• Volume control using finger distance
• Gesture-based media controls
• Multi-hand support

Vehicle Counter

• ROI (Region of Interest) counting
• Object tracking across frames
• Vehicle speed estimation
• Lane-wise analytics
• Traffic statistics dashboard

---

📸 Example Projects

Project	Description
🎵 Hand Gesture Music Controller	Control music with hand gestures
🚗 YOLO Vehicle Counter	Detect and count vehicles in real time

---

🤝 Contributing

Contributions, ideas, and improvements are welcome.

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Open a Pull Request

---

📄 License

This project is licensed under the MIT License.

Feel free to use, modify, and share it for educational and personal projects.

---

⭐ Support

If you found this repository useful, consider giving it a ⭐ Star on GitHub. It helps others discover the project and motivates future improvements.