MindDock

MindDock / sport-vision

Public

基于实时骨骼追踪 & AI 动作识别的羽毛球 / 网球运动智能分析

20
5
100% credibility
Found Feb 26, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Sport Vision is a local web application that analyzes videos of racket sports to track player skeletons, recognize actions like serves and smashes, and visualize biomechanics such as joint angles and movement patterns.

How It Works

1
🔍 Discover Sport Vision

You find this fun tool online that promises to analyze your racket sports videos like badminton or tennis matches.

2
📥 Get it ready

Download the project and run the easy one-click starter to set everything up on your computer.

3
🚀 Launch the viewer

Open your web browser to see a cool, glowing dashboard ready for action.

4
Pick your video
🎬
Try a demo

Select one of the ready example videos to see how it works instantly.

⬆️
Upload yours

Pick a video from your phone or camera of your game.

5
✨ Watch live analysis

As the video plays, glowing skeletons appear on the player, spotting shots like smashes and serves in real time.

6
📊 Check insights

See colorful timelines of every move, heatmaps of footwork, and body stats like joint angles and speeds.

🏆 Master your game

Review the breakdown to understand your technique and improve your racket sports skills.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 20 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is sport-vision?

Sport-vision is a Python tool for analyzing badminton and tennis videos with real-time skeleton tracking and AI action recognition. Upload an MP4 or drop it in the demo folder, hit the one-click run.sh script, and get a web dashboard at localhost:8000 showing pose overlays, biomechanics metrics like joint angles and wrist speed, movement heatmaps, and timelines of actions like serves, smashes, forehands, and lobs. Built on MediaPipe for pose estimation, FastAPI backend, and WebSocket streaming at 20 FPS—all running locally on CPU with zero cloud setup.

Why is it gaining traction?

It stands out with a polished dark neon UI, particle animations, and instant feedback on technique symmetry or body lean, without needing API keys or heavy GPU. Developers dig the rule-based recognition tuned for racket sports, plus easy extension ideas like multi-person tracking. For Sport Vision enthusiasts in Banja Luka, Bih, Hrvatska, Kosovo, MK, outlet stores, Sarajevo, Slovenija, or Srbija, it's a quick way to dissect matches.

Who should use this?

Sports coaches breaking down player form in badminton or tennis drills. Indie devs prototyping vision apps for gym apps or coaching platforms. Researchers in sport vision needing baseline local analysis before scaling to ML models.

Verdict

Grab it for proof-of-concept sports AI—docs are solid, MIT licensed, and setup takes seconds, but with 18 stars and 1.0% credibility score, treat it as an early prototype needing more tests and ML upgrades before production.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.