一种基于C4.5算法的车位识别方法  

Parking plot recognition based on C4.5 algorithm

作　　者：丁兵[1] 吴允平[1] 李彬雅[1] 

机构地区：[1]福建师范大学光电与信息工程学院,福州350007

出　　处：《电子测量技术》2015年第8期64-68,共5页Electronic Measurement Technology

基　　金：教育部科技研究重点(212087);福建省科技厅重大(2011H6009);福建省高校新世纪优秀人才支持计划(JA11037)资助项目

摘　　要：提出了一种基于C4.5算法的车位识别方法。首先通过手机的传感器系统,对车辆行驶轨迹的GPS等数据进行采集,然后利用去除飞点、定位补偿、卡尔曼滤波等方法对采集的数据进行一系列的处理,并进一步采用C4.5算法识别车辆行驶状态信息,最终通过车位识别算法实现对停车位的自动识别。数据处理结果表明,通过上述数据处理过程和方法能够很好的识别潜在的路边车位信息。Proposes a method which is based on C4.5 algorithm to recognize the parking plot.First,we should record the driving traj ectory of the car by the sensors of smart phone in order to obtain the data,such as GPS data,velocity, accelerated velocity ,etc.Then it is necessary that to have the data processed to improve the precision with several means ,such as remove the flying spot ,compensation algorithm ,Kalman Filter and so on .The next step is to get the data further processed through C4.5 algorithm in order to acquire the state of the car .Finally ,the algorithm of recognizing the parking plot is used to make sure whether the spot is a parking plot or not .The results of the processing shows that the method that mentioned above is effective to find the potential roadside parking plot .

关 键 词：C4.5算法 卡尔曼滤波 车位识别 GPS 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]