单目视觉车道线识别算法及其ARM实现

Monocular Camera Machine Vision Lane Recognition Algorithm and Realization on ARM System

首次将ARM技术应用于汽车的自动导航,使用Intel公司生产的基于ARM10的PXA255作为核心处理器,详细讨论了硬件平台与接口的设计。结合单目机器视觉实现了一种基于像素的图像处理算法,该算法能成功地提取道路中的白色车道线,为车辆自动导航提供了准确的图像信息,并将处理时间控制在70 ms以内,取得了很好的实时效果。同时,建立了一种简单有效的透视投影模型,实现图像坐标系与摄像机坐标系之间的映射,不仅为图像处理算法提供了有效的约束条件,而且获取了道路的深度信息。在沪宁高速南京段的实验结果表明该算法的实时性与精确性已基本满足安全行驶要求。

The ARM technology is used to simulate vehicle automatic navigation for the first time. PXA255, an Intel XScale microchip based on ARM10, is used as the core processor. The design of the hardware platform and the interface is discussed. An image processing algorithm based on the pixel and the monocular vision is developed to recognize the white strip of a highway and to provide valid image information for vehicle automatic navigation. Experiments show that runtime of the algorithm is less than 70 ms. A transform mapping model is established to limit the region of the algorithm and obtain the depth information of the road. Experiments are conducted on Hu-ning highway Nanjing section. Results show that the algorithm is fast and accurate for the safety of driving.