基于道路特征信息的视觉车道模型研究
Visual Lane Model Based on Road Characteristic Information
-
摘要: 为有效进行道路设计评价以及分析道路环境影响驾驶员行为的机理,提出一个视觉车道模型,用于描述驾驶员眼中的道路特征信息. 视觉车道模型基于Catmull-Rom样条曲线建立,在CAD平台下自主开发了视觉车道模型软件,能自动进行设计道路视觉图像拟合;经大量实验分析表明,该模型能拟合任意形状的道路线形,并提取相关参数. 对视觉车道模型与道路三维几何信息的关系进行统计分析,是进行道路设计一致性评价的基础.Abstract: A novel visual lane model (VLM), which can be used to describe the road characteristic information in driver's perception, is proposed to evaluate road design and analyze the mechanism of driving behavior. The VLM is constructed based on Catmull-Rom spline. Then a VLM software which can conduct road visual image fitting automatically is developed on Auto CAD platform. Experimental results show that the proposed model can fit road alignment with arbitrary shape, and collect related road characteristic information.
下载: