根据地形坡度旋转的车载激光点云形态学滤波

发布时间：2018-04-21 18:30

本文选题：开运算 + 形态学滤波　；参考：《遥感信息》2017年01期

【摘要】：针对传统的基于开运算的形态学滤波方法在坡度连续变化较大的区域无法有效滤除非地面点的问题,提出了一种可适应斜坡地形的滤波新方法。首先对全区域进行分块,构建地面不规则三角网确定各分块区域的地形坡度和方向;然后根据地形坡度和方向对各分块区域内的数据进行旋转,使地形坡度接近于水平,从而更适用于基于开运算的形态学滤波方法;最后对点云数据执行多尺度形态学开运算确定地面点。分别选择平坦地形和斜坡地形的车载激光点云数据进行了实验,两类点云数据均包含植被、建筑、裸地等,结果表明,提出的方法能够有效滤除各种类型地物,并且对斜坡地形具有较好的滤波效果。
[Abstract]:Aiming at the problem that the traditional morphological filtering method based on open operation can not effectively filter except the ground points in areas with large continuous slope changes, a new filtering method for slope topography is proposed in this paper. Firstly, the terrain gradient and direction of each block area are determined by constructing the ground irregular triangle network, and then the data in each block area are rotated according to the terrain slope and direction, so that the terrain slope is close to the level. So it is more suitable for morphological filtering method based on open operation. Finally, multi-scale morphological open operation for point cloud data is performed to determine the ground point. The experimental results show that the two kinds of point cloud data include vegetation, building, bare land and so on. The results show that the proposed method can filter all kinds of ground objects effectively. And the slope terrain has better filtering effect.
【作者单位】：山东科技大学海洋工程研究院;海岛(礁)测绘技术国家测绘地理信息局重点实验室;山东科技大学测绘科学与工程学院;
【基金】：测绘公益性行业科研专项经费资助项目(201512034) 海洋公益性行业科研专项经费资助项目(201305034) 山东省高等学校科技计划项目(J13LH04) 海岛(礁)测绘技术国家测绘地理信息局重点实验室资助项目(2015A01) 山东科技大学教育教学研究“群星计划”项目(qx2013307);山东科技大学研究生教育创新计划资助项目(KDYC15004)
【分类号】：P237

【参考文献】