摘要 : Thermal comfort during sleep is essential for both sleep quality and human healthwhile sleeping. There are currently few effective contactless methods for detectingthe sleep thermal comfort at any time of day or night. In this pap... 展开 Thermal comfort during sleep is essential for both sleep quality and human healthwhile sleeping. There are currently few effective contactless methods for detectingthe sleep thermal comfort at any time of day or night. In this paper, a vision-baseddetection approach for human thermal comfort while sleeping was proposed, which isintended to avoid overcooling/overheating supply, meet the thermal comfort needs ofhuman sleep, and improve human sleep quality and health. Based on 438 valid questionnairesurveys, 10 types of thermal comfort sleep postures were summarized. Byusing a large number of data captured, a fundamental framework of detection algorithmwas constructed to detect human sleeping postures, and corresponding weightingmodel was established. A total of 2.65 million frames of posture data in naturalsleep status were collected, and thermal comfort-relatedsleep postures dataset wascreated. Finally, the robustness and effectiveness of the proposed algorithm werevalidated. The validation results show that the sleeping posture and human skeletonkeypoints can be used for estimating sleeping thermal comfort, and the the quilt coveragearea can be fused to improve the detection accuracy. 收起