LPCSB: A Device to Distinguish Between Natural and Artificial Light
Sarris, Alexander, Computer Engineering - School of Engineering and Applied Science, University of Virginia
Campbell, Brad, EN-Comp Science Dept, University of Virginia
Studies have shown that natural lighting (sunlight) is very beneficial for office workers' productivity and morale. While many new office spaces are being designed to emphasize the use of natural lighting over artificial lighting, it is difficult to determine whether sunlight actually illuminates the majority of such a workspace during the day. We present the initial results of our work in developing a new small, low-power, and wireless system that can distinguish between natural and artificial light. Previous works focus on analyzing measured light in the frequency domain. Our system analyzes light in the visible spectrum, utilizing an off-the-shelf RGB photodiode array that measures incoming light in conjunction with an algorithm that classifies the light into one of five categories - "Incandescent", "Fluorescent", "LED", "Sunlight", and "Unknown" - based on various characteristics derived from analyzing the measured components of the raw data in the visible spectrum of light (red, green, and blue). Initial testing has shown the classification algorithm to be very effective. The algorithm can correctly identify all of our tested artificial light sources with over 80% accuracy and natural light with over 75% accuracy.
MS (Master of Science)
Color Sensing, Light Type Identification, Visible Spectrum, Bluetooth Low Energy, Natural Light, Artificial Light, LPCSB, Low Power Color Sensing Board
English
All rights reserved (no additional license for public reuse)
2020/04/22