论文部分内容阅读
针对目前高分辨率遥感影像的道路自动提取算法研究中的不足,该文提出了一种基于并行角度纹理特征的半自动道路提取算法:用户输入完成道路中心线上的起始点、道路方向、道路宽度等初始化工作,利用并行角度纹理特征获取道路前进方向,用抛物线参数方程构建道路轨迹模型来预测道路轨迹点,使用角度纹理特征值构建的紧质度系数和抛物线的曲率变化来约束道路轨迹点,验证失败则转入手工跟踪;往复执行以提取道路中心线。试验证明,本算法是一种稳健的道路半自动提取算法。
Aiming at the shortcomings of the current road automatic extraction algorithm for high resolution remote sensing images, this paper proposes a semi-automatic road extraction algorithm based on parallel angle texture features: the user inputs the starting point of the road center line, the direction of the road, the width of the road The pavement parameter is used to construct the trajectory model to predict the trajectory points. The constrains of trajectory points are constrained by the tightness coefficient and parabola curvature of the angle texture eigenvalue, The verification fails into manual tracking; reciprocating execution to extract the centerline of the road. Experiments show that this algorithm is a robust semi-automatic road extraction algorithm.