The Design of Vehicle Headlights Switching System Based on Support Vector Machine(SVM)

摘要

This article aims at proposing a design of an intelligent switching system for vehicle headlights based on Support Vector Machine (SVM) due to the fact that the improper use of the vehicle headlights leads to some of the traffic accidents. Firstly, the system uses the AR0130 image sensor to capture road images. Second, the system classifies the image to identify whether there is a streetlight, whether there is an oncoming vehicle and whether it is too close to the front vehicle in the absence of a streetlight with the application of Raspberry Pi 3 based on SVM. Then the corresponding correct indication of the headlights switching is given by the control of relays, reducing the driver's distraction greatly. Experimental results prove that the system has high practicality and accuracy for the switching of vehicle headlights.