论文部分内容阅读
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.
Intelligent vehicle research is a hot research field in the world and an important direction of computer vision application.Visual system is a key part of intelligent vehicles, of which road recognition and tracking algorithms are the core of intelligent vehicles.This paper uses robust feature space analysis Method - mean shift method, the road color information is clustered, and the clustering result is combined with the reference area method to realize the road recognition, and the travelable area of the smart vehicle is obtained. The experimental results show that the algorithm has strong robustness and autonomy Sexual and parallelism, especially suitable for real-time image processing systems, such as intelligent vehicle vision systems, mobile robots visual navigation section.