基于机器学习的车辆路面类型识别技术研究

摘要

当车辆在各种不同的路面上行驶时,获知路面类型信息将有助于提高乘车人的安全性和舒适性,不同的路面类型将对车辆的加速、制动及操控等驾驶策略产生影响.基于机器学习的基本原理,提出一种使用加速度传感器和相机特征数据融合对路面类型进行分类的方法,并与单独使用其中一种传感器进行了比较.使用垂直加速度和车速数据并利用车辆动态模型还原路面轮廓,进而完成特征提取和路面类型分类;对相机采集的路面图像数据进行特征提取和分类;将两类传感器的数据特征进行融合,完成路面类型识别任务.实验结果表明:使用两种传感器数据特征融合的方法,不但识别精度有所提高,而且其可靠性和适应性也都优于单独使用加速度数据或路面图像数据.%The acquisition of information about the road terrain helps to improve the passengers'safety and comfort when a vehicle runs on different road terrains.Different road terrains have significant impacts on the driving acceleration,braking and manipulation of vehicle.A machine learning-based recognition method is proposed,which is to recognize the road terrain by fusing the feature data from accelerometer and camera.The road profile is estimated by using acceleration and vehicle speed data.The spatial features are extracted from the road profile for terrain classification.The texture features extracted from terrain images captured by a camera are used for the same classification task.And the task of recognition of road terrain is accomplished by fusing the data features from two sensor data sets.The experimental results show that the proposed method is used to improve the accuracy of road terrain recognition,and the reliability and comfort of passengers in vechile.