基于BP神经网络和树莓派的自主驾驶车设计

Autonomous Driving Vehicle Design Based on BP Neural Network and Raspberry Pi

随着传感器技术和人工智能技术的快速发展,无人驾驶车研究已成为当前机器人领域的研究热点.无人驾驶车利用树莓派作为主控核心设计,以CMOS摄像头作为道路和红绿灯图像采集模块,以直流减速电机和L298N芯片作为驱动模块,根据道路情况利用BP神经网络算法实现车道线的识别,并通过AdaBoost级联分类器对道路红绿灯进行判别,从而实现道路和红绿灯的自动识别和自主驾驶.

The accelerated advancement of sensor technology and artificial intelligence has prompted a heightened emphasis on research into autonomous vehicles,which is currently one of the most prominent topics in the field of robotics.This paper presents the design and implementation of an autonomous driving car,in which the Raspberry Pi is employed as the principal control core.A CMOS camera is utilized as the image acquisition module for the road and traffic lights,while DC decelerator motors and an L298N chip constitute the drive module.The BP neural network is designed to recognize lane lines in accordance with the prevailing road conditions,while the AdaBoost cascade classifier is employed to distinguish the road traffic lights,thereby enabling the automatic recognition between roads and traffic lights and the realization of autonomous driving.