基于摄像头的道路识别及赛车控制算法研究  

Research on road styles identifying technique of camera and the control algorithm of vehicle

作　　者：丁炜[1] 吕梁[2] 齐冬莲[2] 

机构地区：[1]兰州石化职业技术学院电子电气工程系,甘肃兰州730060 [2]浙江大学电气工程学院,浙江杭州310027

出　　处：《电子设计工程》2011年第3期178-180,183,共4页Electronic Design Engineering

基　　金：国家自然科学基金资助项目(60702023);甘肃省教育厅资助项目(0915B-1)

摘　　要：为了提高智能车对道路的预判能力,探讨了摄像头在智能车道路识别中的应用,提出一种通用的控制算法。在准确采集图像的基础上,通过图像还原技术获得赛道的真实信息,利用临近搜索法对有效道路信息进行快速提取,计算出赛道黑线的走向趋势及赛车当前位置,对赛道进行合理地分类和优化,根据不同的赛道选不同频率的PWM波,极大地减小加速和减速的时间,采用Bang-Bang控制算法使速度达到理想值。To improve the smart car＇s ability of pre-judgment to the road, this paper studied the application of identifying the path by using the camera, put out a kind of general control algorithm. On sampling images accurately,it obtained the true information of the road by using image restoration techniques and picked u＇p the effective information of the road by using neighbor extraction method. So the trend of black-line and the current position the car against the center of black-line was figured out,and the information for road styles was classified and optimized. Experiments show that the different frequency of PWM can greatly reduce speed up and down the time,a Bang-Bang control algorithms gets the desired speed value.

关 键 词：图像采集 临近搜索 转向控制 速度控制 Bang—Bang控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]