【摘要】 智能车辆研究是目前世界各国学者研究的热点,也是计算机视觉应用的一个重要方向.视觉系统是智能车辆的关键部分,其中道路识别和跟踪算法是智能车辆的核心部分.本文用鲁棒特征空间分析方法—均值移动法,对道路颜色信息进行聚类,聚类结果结合参考区域法实现道路的识别,从而得到智能车辆的可行驶区域,实验结果表明,该算法具有较强的鲁棒性、自主性和并行性,特别适用于实时图像处理系统,如智能车辆视觉系统、移动机器人视觉导航部分.更多还原

【Abstract】 Intelligent vehicle is an active research domain over the world, and is also an important research direction of computer vision. Vision system is the key part of intelligent vehicle, in which road recognition and following arc the core modules. In this paper, the robust spatial analysis method, mean shift, is adopted to cluster color information of road, and the road is recognizd using a reference region. Then the region that vehicle can run can be got. Experimental results indicate that the method has good properties of robustness, autonomy and parallelism, and it is suitable for real time image processing systems, such as the computer vision system of intelligent vehicles, navigation part of mobile robots.更多还原