This project is an innovative web application designed to provide real-time pose detection for users, specifically geared toward yoga and fitness enthusiasts. The application was built using the Django framework for its robust backend. It serves as a platform for users to log in, practice yoga poses, and receive immediate feedback on their form through live video feeds.
Key Features:
- Pose Detection and Feedback: Leveraging MediaPipe’s pose estimation, the application uses a webcam feed to detect body landmarks and classify yoga poses such as the Tree Pose, Warrior II Pose, and T Pose. Users receive real-time corrections on their forms.
- Automated Screenshots: The system takes screenshots when detecting the correct pose, allowing users to review their progress.
- Live Video Feed: The webcam feed is streamed in real-time, and the application uses OpenCV to process and display feedback dynamically.
- User Authentication: The app includes secure user authentication using Django’s built-in authentication features. It supports user sign-up, login, and account activation via email verification.
- Responsive Design: The front-end pages, such as home, blog, and contact, are clean and user-friendly, ensuring a smooth user experience.
Tech Used:
- Django: Backend framework for building the web application and handling user authentication.
- OpenCV: For capturing and processing live video input from the user’s webcam.
- MediaPipe: For real-time pose detection and body landmark recognition.
- PyAutoGUI: To capture screenshots when the correct pose is detected.
- Threading: Ensures smooth video streaming and uninterrupted performance.