基于线性模型的导航路径图像检测算法研究

Study on Linear Model Based Image Recognition Algorithm of Navigation Route

对智能车辆视野内的车道采用线性模型进行拟合 ,分析线性模型拟合车道的误差 ,提出减小误差的方法 ,并分析户外环境中几种典型光照变化和阴影对智能车导航的影响 ,改进基于线性车道模型的边缘检测算法。实验证明 :在JUTIV II上采用线性模型可提高导航路径识别的准确性 ,减少数据处理量和获得较高计算机处理频率。

In this paper, a linear model is used to describe the lane in the visual field of the intelligent vehicle. The error of the linear model is analyzed and a method is used to reduce the error. Several typical illumination changes are analyzed and an enhanced algorithm is proposed to overcome the disturbance. From the field experiment, it is found that the route Recognition algorithm based on the linear model has high reliability, reduces the calculation work greatly and increases the control frequency.