Binocular stereo vision based illuminance measurement used for intelligent lighting with LED

Abstract Intelligent lighting is responsible for providing an appropriate illuminance and balancing the environment and resources. An appropriate illuminance can improve visual effect of machine recognition, while it is always difficult to be measured and controlled. An illuminance measurement method by using depth measurement technology of binocular stereo vision is proposed in this study. Illuminance analysis models of single LED and combined LEDs are established, and theoretical calculation equations for analyzing the relationship between power, illuminance, and depth are proposed. The illuminance distribution of single LED and combined LEDs are measured. The difference between single LED and combined LEDs are obtained in terms of illuminance, power, energy efficiency, effective illuminance area, etc. The results show that the illuminance is proportional to power of LED consumed and inversely proportional to the square of depth of image. The higher the energy conversion efficiency under the high illuminance for both single LED and combined LEDs. The depth of image is measured by binocular stereo vision technology, which has a good measurement performance. The binocular stereo vision based illuminance measurement method proposed in this study is efficient, reliable and robust, which is in good agreement with the experimental results, and the relative error is less than 3.8 %. The illuminance measurement is realized without increasing the equipment and image acquisition workload, which lays the foundation for dynamic monitoring and controlling of illuminance. The research results can also be applied to related fields as well, such as the intelligent lighting of street lamp, high beam in automatic driving and supplement lamps in video monitoring.