Research on a high-altitude falling monitoring and early warning system
This paper addresses the issue of children falling from heights,by proposing a monitoring and warning system that utilizes target detection and ranging technologies to mitigate high-altitude falls.The sys-tem has improved upon the existing challenges posed by large parameter sizes in the YOLO object detection algorithms,and introduced a new hazard distance measurement algorithm based on monocular camera ranging techniques.The experimental results indicate that the improved object detection algorithm,while maintaining similar performance in other parameters,reduces the model size by 1.99 MB and the number of parameters by 1.039 billion compared to the original YOLOv8n model.The newly designed hazard distance measure-ment algorithm achieved a mean lateral error of 2.29 cm,a mean longitudinal error of 3.16 cm and a hazard monitoring warning accuracy of 92.00%across 100 data sets in four different scenarios.In summary,the high-altitude falling detection and early warning system presented in this paper exhibits high accuracy and ro-bust practicality,making it highly valuable for societal applications.
Falling from a heightChild safety monitoringTarget detectionDistance measurementHaz-ard warning
国家科技期刊平台
NSTL
万方数据
