Patents
Proprietary AI Patent Portfolio
AISION holds two patented AI Motion Analysis technologies that integrate OpenCV and Human Pose Estimation.
Built on advanced human posture and motion analysis, our technology is designed to expand across a wide range of applications, including sports, gaming, entertainment, and digital healthcare.
<AI Motion Analysis Flow>
Pose Data Processing Engine
- Google MediaPipe Human Pose Detection
- Represents 33 human landmarks using more than 150 motion parameters
- Analyzes large-scale motion data to build optimized movement and posture models
- Modularized into more than 50 exercise and posture models
AI Posture Coaching
- Deep Learning trained on posture and movement data to establish reference models
- Compares and tracks joint angles, distances, and positions against reference data in real time
- Provides real-time posture correction through visual analytics and voice coaching
01
AI Personal Trainer System
AI Motion Analysis has been pre-trained on more than 150,000 human poses using Google's Pose Detection API to accurately recognize human posture and movement.The system estimates 33 three-dimensional human landmarks and utilizes a trained posture model to provide real-time voice-guided coaching for exercise performance.
02
Doctor–Patient AI Exercise Prescription Platform
AI Motion Analysis supports musculoskeletal rehabilitation, orthopedics, neurosurgery, and physical medicine & rehabilitation by enabling healthcare professionals to deliver personalized rehabilitation and exercise prescription programs.
The platform is particularly effective for Tele-Rehab and patients with limited mobility.
① Healthcare professionals create and assign personalized exercise prescriptions, including exercise type, training method, duration, repetitions, and rest intervals.
② Patients perform their prescribed exercise program through the AI platform.
③ Exercise results, training duration, and personalized performance data are automatically transmitted to healthcare professionals for monitoring and evaluation.
