การพฒนาระบบตรวจจบการลมของผสงอายดวยคอมพวเตอรวทศน

Elderly individuals are particularly susceptible to fall-related accidents, especially those living alone. This study proposes the development of a computer vision-based fall detection system designed to enhance personal safety through real-time monitoring. The system integrates indoor video surveillance using OpenCV and processes the captured frames via Media Pipe to extract and analyze human postures based on 33 skeletal landmark points. Fall detection is achieved by identifying abrupt downward movements toward the ground, measured in terms of both direction and duration. Experimental simulations covering various fall scenarios—namely forward, backward, leftward, and rightward—demonstrated that the system achieved an average detection accuracy of 90%. These findings underscore the system’s potential to reduce the risk of injury among elderly individuals by enabling timely intervention in the event of a fall. It is anticipated that the outcomes of this research will contribute to the advancement of intelligent fall detection technologies and provide a foundation for future developments in elderly care and health monitoring systems.