Objective:
To develop a method that can continuously monitor duration of students outdoor activities for a long time, so as to provide data support for the relationship between outdoor activity duration and students health.
Methods:
From April 28 to July 6, 2022, 1 168 students from a primary school in Shenzhen were selected. Fixed cameras were placed on the top of school classrooms, corridors and critical paths were used to identify student data and duration in the picture. And AI, cloud computing and other methods were used to measure students-non-classroom time instead of outdoor activity time in school.
Results:
The average length of time spend on outdoor activities in school of the 24 pilot classes were 67.6-113.0 min, and showed a downward trend by grade ( F =42.74, P <0.05). The duration of outdoor activities among students was negatively correlated with the detection rate of myopia and overweight( r =-0.74, -0.45, P <0.05).
Conclusion
The data on outdoor activity time calculated by AI image recognition is basically in line with the reality. Monitoring students outdoor activity time at school through AI image recognition is feasible and popularized.