A comprehensive functional fitness assessment system leveraging openpose technology for elderly individuals

With the rapid growth of the aging population, assessing the physical fitness of elderly individuals has become increasingly important for maintaining independent living and improving quality of life. Traditional functional fitness assessments are typically conducted through manual observation and recording, which may introduce human errors and lack real-time feedback on movement correctness. To address these limitations, this study proposes a computer vision–based functional fitness assessment system designed for elderly individuals. The proposed system utilizes a standard webcam to capture motion images and employs the OpenPose framework to detect human body joints and construct skeletal representations of the participants. Based on the extracted joint coordinates, geometric analysis methods including joint angle calculation, distance measurement, and slope analysis are applied to evaluate whether the performed movements satisfy predefined functional fitness testing criteria. Four functional fitness test items were implemented in the system, including bicep arm curl, chair stand, step-in-place, and single-leg stance. Experimental results demonstrate that the proposed system can effectively recognize functional fitness actions with high accuracy. The system achieved recognition accuracies of 95.68% for bicep arm curl, 91.67% for chair stand, 91.11% for step-in-place, and 92.11% for single-leg stance, respectively. The results indicate that the proposed system provides a practical and accessible solution for automated functional fitness assessment. By utilizing a low-cost webcam and pose estimation technology, the system has strong potential for applications in elderly health monitoring, rehabilitation assessment, and community-based fitness evaluation.