Image Recognition and Extraction of Students' Human Motion Features Based on Graph Neural Network

Area of Science:

• Sports Medicine
• Computer Vision
• Biomedical Engineering

Background:

• Teenagers' physical health and subhealth behaviors are significant concerns.
• Schools are mandated to increase student exercise frequency and duration.
• Understanding physical changes and injury impacts during student sports is crucial.

Purpose of the Study:

• To propose a student human motion feature image recognition system based on graph neural networks (GNNs).
• To analyze physical changes during sports activities and the health impacts of sports injuries.
• To investigate factors contributing to physical subhealth in classroom settings.

Main Methods:

• Combined image recognition technology with GNN management for target detection and tracking.
• Utilized image recognition to analyze student movements and identify potential injuries.
• Employed GNNs to manage detection and tracking results for visual data representation.

Main Results:

• Image recognition accurately analyzed cervical spine and sports injuries during student activities.
• The GNN approach provided accurate experimental data for student physical health analysis.
• Visual expression of student physical health test data was achieved through GNNs.

Conclusions:

• The proposed GNN-based image recognition system effectively analyzes student sports injuries.
• Accurate data collection supports analysis of student physical health and subhealth factors.
• This technology offers valuable insights for promoting healthy student development and preventing injuries.