倾斜影像中提取高精度DEM的方法研究

针对传统航空影像获取的DSM在立面及局部地面、建筑物屋顶空间信息的不足，获取高精度DEM较为困难的问题，提出了基于倾斜影像提取高精度DEM的方法。首先对倾斜影像获取的点云DSM结构进行分析，得出了DSM具有几何约束特点，能够在城区很好地区分地面点和地物点；然后指出对DSM滤波处理是获取高精度数字高程模型（DEM）的关键技术，提出了基于法向量差值区域生长分割TIN的滤波方法；最后选取吉林省敦化市的倾斜影像数据进行了滤波试验和算法验证。试验结果表明，该方法能够快速、有效地滤除不同尺寸的建筑物、植被和其他地物，获取高精度DEM。

Study on the Method of Extracting High Precision DEM from Oblique Image

For the DSM of the traditional aerial images acquisition, the lack information of the facade,the local floor and the roof of the building, the problem which difficult to obtain high precision DEM, the method of extracting high precision DEM from oblique image is proposed in this paper. Firstly, the structure of cloud point of oblique image acquisition is analyzed, the DSM has a geometric constraint characteristics which can well distinguish between ground points and ground feature points in the city. Then it points out that the filte-ring is the key technology to obtain the high precision digital elevation model (DEM), and proposes a filtering method based on region growing segmentation TIN.Finally, the filtering experiment and algorithm validation are carried out by the oblique image data of Dunhua City, Jilin Province.The experimental results show that the method can quickly and effectively remove the different sizes of buildings, vegetation and other objects, and obtain high precision DEM.