Hand Gesture Recognition System

A real-time hand gesture recognition system developed using Python, TensorFlow, MediaPipe, and OpenCV.

Overview

This project recognizes hand gestures from a webcam feed. It uses MediaPipe to detect hand landmarks and a TensorFlow model to classify gestures in real time.

Features

• Real-time gesture detection
• Hand landmark extraction using MediaPipe
• TensorFlow-based gesture classification
• Custom gesture data collection
• Model training and evaluation
• Live prediction through webcam

Screenshots

Peace Gesture

Palm Gesture

Fist Gesture

OK Gesture

Call_me Gesture

Model Evaluation

Technologies Used

• Python
• TensorFlow
• MediaPipe
• OpenCV
• NumPy
• Scikit-Learn
• Matplotlib
• Seaborn

Project Structure

gesture_recognition_system.py
images/
models/
gesture_data/
README.md
requirements.txt

Installation

Clone the repository:

git clone https://github.com/npatil09/gesture-recognition-tensorflow.git

Install the required libraries:

pip install -r requirements.txt

Usage

Collect gesture samples:

python gesture_recognition_system.py --mode collect --gesture peace

Train the model:

python gesture_recognition_system.py --mode train

Start real-time detection:

python gesture_recognition_system.py --mode detect

Supported Gestures

• Fist
• Palm
• Peace
• OK
• call_me

Future Improvements

• Add more custom gestures
• Improving model accuracy
• Gesture-controlled system actions
• Presentation control using gestures