Research On Key Techniques Of Intelligent Lighting System Based On Video Recognition

Intelligent lighting is an important direction of future lighting technology development, which can improve the quality of the lighting environment and satisfy people’s personalized lighting environment requirements. The most commonly used methods include the controls based on time, illuminancy and luminous flux, etc. In regards of the above inadequacies, this paper has proposed innovatively the intelligent lighting control system based on video identification, for which the key algorithms has been designed, and the algorithm has been validated.The paper starts off briefly introducing the intelligent lighting system and analyses the research status and development trend of intelligent lighting domestically and abroad. In regards of the inadequacies of the existing lighting control system, this paper has proposed the intelligent lighting control system based on video identification, for which the key algorithms has also been designed. The target detection by combining a three-frame difference method and mixed Gaussian background modelling, and moving target tracking by combining Kalman filtering and Cam Shift algorithm, the intelligent lighting system designed in this project has the capability of real-time dynamic adjustment according to indoor figures’ changes in movements, positions, time, which has improved the interactivity between lighting system, users and the environment.On this basis, the paper has realized the algorithm based on Beagle Bone Black hardware platform, and has provided detailed explanations to the methods of building embedded Linux cross complier tool chain based on PTXdist tools, Linux kernel migration and Open CV migration.