Smart Lighting System for Commercial Buildings using Digital Camera

Artificial lighting systems play a major role in en-ergy consumption of commercial buildings. An energy conscious automatic lighting design can help in bringing down the energy utilization and can also boost the human convenience level. A smart LED lighting system that uses a digital camera as an input sensor which detects occupancy is proposed in this paper. With the help of the input from digital cameras, the need for a separate network of conventional IR sensors to detect the occupancy in the room is eliminated. Different object detection techniques are studied along with their constraints in real-time demonstration and SSD- Mobilenet architecture is chosen as the object detection technique considering the trade-off required between speed of operation and accuracy. Finally, the proposed model is implemented on a Raspberry Pi 3b+ model with LEDs to mimic the real-time operation of the lights in a hall.