Computer Vision Experiments Lab

Interactive real-time computer vision experiments built with Python. The application demonstrates face tracking, hand tracking, gesture-based control, augmented reality overlays, and air drawing using a webcam.

---

Features

Face Landmark Detection

Real-time facial landmark tracking using MediaPipe. The system detects key facial points and visualizes them directly on the video stream.

Hand Landmark Detection

Tracks finger positions and hand structure with high accuracy.

AR Hand Effect

A transparent Doctor Strange–style visual effect follows the detected hand in real time, creating a simple augmented reality interaction.

Gesture-Based Mouse Control

Control your computer using hand gestures:

• Move cursor
• Click actions
• Touchless interaction

Virtual Air Drawing

Turn your index finger into a digital brush:

• Draw in real time
• Clear the canvas
• Smooth motion tracking

---

Tech Stack

• Python
• Streamlit
• MediaPipe
• OpenCV
• PyAutoGUI

---

Project Structure

Computer-Vision-Experiments-Lab
│
├── src
│   ├── main.py
│   ├── facelandmark.py
│   ├── hand_landmark_mask.py
│   ├── mouse_control.py
│   └── drawing_mode.py
│
├── models
│   ├── hand_landmarker.task
│   └── face_landmarker.task
│
├── assets
│   ├── dr_strange_hand.png
│   └── demo.gif
│
├── requirements.txt
└── README.md

---

Installation

git clone https://github.com/yourusername/Computer-Vision-Experiments-Lab.git
cd Computer-Vision-Experiments-Lab
pip install -r requirements.txt

---

Run the Application

streamlit run src/main.py

Open the local Streamlit address shown in the terminal.

Allow camera access when prompted.

---

Models

Used Models

Hand Landmarker

Face Landmarker

---

How to Use

1. Launch the application.
2. Select a mode from the interface.
3. Position yourself in front of the webcam.
4. Try different gestures.

Available modes:

• Face Landmark Detection
• Hand Landmark Mask
• Hand Mouse Control
• Drawing Mode

---

Future Improvements

• Gesture shortcuts
• Multiple brush colors
• Adjustable brush size
• Performance optimization
• Additional AR effects

---

Author

YamenRM AI Engineering Student