Real-time smart lighting control using human motion tracking from depth camera

A smart lighting system provides automatic control of lighting illumination and color temperature for high quality of life as well as energy savings in smart cities. A real-time activity understanding system with accurate human location estimation under varying illumination is required for smart lighting control, since comfortable lighting conditions vary according to human activities. This paper presents a real-time smart lighting control system using human location estimation based on inverse-perspective mapping of depth map images and activity estimation from location, heading direction, and height estimation of the moving person from multiple depth cameras. Lighting control based on estimated proximity to the specific activity area, distance to the target lighting area, and heading direction of the person provides an automatic activity-dependent lighting environment as well as energy savings. We implemented several activity modes such as study mode, dialog mode, and watching TV mode, and applied the proposed lighting control system to a living room lighting control with known furniture, electronics, and lighting locations using multiple Kinect depth cameras. The proposed model is based on localized proximity-based lighting control and can be extended to a more general lighting control by combining with global lighting control schemes.