边缘检测算子LiDAR河流数据提取的研究

由于LiDAR数据具有离散、不连续的特点,在空间上并没有明确的函数关系,现有的LiDAR数据地形地物的提取算法存在依赖辅助数据和算法只针对建筑物等明显地物提取的问题。河流的数据存在形状不规则、边缘不明显的特征,自然形成的原因导致其坡度信息多变,常见的数据提取算法难以解决这类问题,Robert边缘检测算子对边缘信息更为敏感,实验证明,与Sobel和Prewitt算子对比,Robert算子更适用于提取较高精度河流地物边缘并且有较好的精确性和适用性,可获得较高精度作为准确的目标信息。

Study on Getting Edge Detection Operator LiDAR River Data

For LiDAR data has the disperse and discontinuous characteristic and hasn't clear function rela-tionship in space,the problem is that the pick-up algorithm of present LiDAR data from terrain and surface features depends largely on auxiliary data and algorithm is only for surface features which has obvious buildings.The data about river has out-of-shape and unconspicuous edge feature,and nature reason causes that gradient information is changeful.The common data pick-up algorithm can't solve this problem.Robert edge detection operator is more sensitive to edge information,then,the experiment proves that,compared with Sobel and Prewitt operator,Robert operator is more suitable to pick up high precision river land fea-ture edge with better precision and applicability,it can get more precisely target information.