Skip to content

An AI-powered tool that enhances martial arts practice by analyzing and grading movements. Using OpenCV, it compares user performance with tutorial videos, provides real-time feedback, and delivers personalized scores to help learners improve at their own pace.

License

Notifications You must be signed in to change notification settings

hasan-mustafaa/MartialVision

Repository files navigation

MartialVision

It's a platform designed to enhance martial arts practice by analyzing and grading movements. The system uses computer vision to compare a user's performance with tutorial videos and provides detailed feedback to help them master their skills.

Key Features:

Unlimited Capabilities:

  • Users can upload any video demonstrating any form of martial art.

Tutorial Playback:

  • Plays segmented videos of martial arts movements for users to follow.

User Video Capture:

  • Records the user's performance of the selected martial arts section.

AI-Powered Grading:

  • Uses OpenCV to analyze and compare the user's movements with the tutorial.
  • Splices videos into comparable sections for accurate grading.
  • Scores performance based on posture, precision, and timing.

Cumulative Scoring and Feedback

  • Highlights areas of improvement using feedback.

How It Works

Step 1: Choose a Tutorial

  • Upload a specific martial arts technique or movement to practice.

Step 2: Record Performance

  • Perform the action while the system records your movements via webcam or phone camera.

Step 3: AI Analysis

  • OpenCV tracks body motion and compares it with the tutorial video.
  • Identifies deviations in joint angles, timing, and accuracy.

Step 4: Get Your Score

  • Suggestions for improvement are displayed.

About

An AI-powered tool that enhances martial arts practice by analyzing and grading movements. Using OpenCV, it compares user performance with tutorial videos, provides real-time feedback, and delivers personalized scores to help learners improve at their own pace.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •  

Languages