Review of vision-based occupant information sensing systems for occupant-centric control

摘要

Vision-based (camera-based) systems, which can effectively sense occupant information, have garnered attention as a core technology in the Fourth Industrial Revolution. A detailed understanding of vision-based sensing systems is required to detect occupant information based on vision and use it for occupant-centric control. Therefore, in this study, we performed a comprehensive and structural literature review of vision-based occupant information systems. The contributions of this review can be summarized in the following six points: (1) a fivetier taxonomy of vision-based occupant information is proposed, (2) a systematic summary of vision-based occupant information is presented, (3) the quantitative and qualitative performance of sensing systems is reviewed, (4) an analysis of the applicability of deep-learning-based computer vision techniques is presented, (5) a summary of privacy-preserving techniques is included, and (6) a summary of vision-based control strategies and energy saving potential analysis is provided. The analysis in this review is an important contribution toward addressing the challenges in the field of research.