A Novel Intelligent Lighting Control System Based on Object Detection

This novel intelligent lighting control system is designed for the smart lights on the car at night, using computer vision to collect road information in the night and give the fruits to the controller to control the car's lights. Thus, the night lights can be controlled intelligently. The system consists of a night road detection and image acquisition module, image feature extraction and object detection module, and light status control module. Firstly, the road information is obtained from a camera, and then the image is processed and detected. According to vehicles' characteristics at night, the feature vectors of a headlight, taillight, street lights, and other bright spots are extracted. The extracted feature vectors are imported into a convolution neural network for training and verification of the training set and test set. After obtaining the pavement information results, the light status control module determines the intelligent light's strategy. The performance of the intelligent lighting control system is verified.