Quantitative Evaluation and Optimization of Museum Fatigue Using Computer Vision Human Pose Estimation

Museums are key institutions for cultural communication and public education, and their operating concept is shifting from exhibit-centered to experience-centered. As expectations for exhibition experience rise, museum fatigue has become a major constraint on visitors. Existing studies rely on questionnaires and other subjective measures, which makes it difficult to locate fatigue in specific spaces. At the same time, body pose detection and fatigue recognition techniques remain hard to apply in museums because of complex spatial configurations and dense visitor flows. Effective methods for quantifying and mitigating museum fatigue are still lacking. This study proposes a contact-free sensing scheme based on computer vision and builds a coupled analytical framework with three stages: Human Pose Estimation (HPE) for visitor posture detection, fatigue assessment, and fatigue mitigation. A Fatigue Index (FI) quantifies bodily fatigue. Applying this index to the exhibition space in both the baseline and adjusted configurations guides the formulation of mitigation strategies and shows a consistent reduction in FI, which indicates that the adopted measures are effective. The proposed approach establishes a complete frame from fatigue quantification to fatigue mitigation, supports evaluation of exhibition space design, and provides theoretical and methodological support for future improvements to museum experience.