基于航空影像和地形自适应滤波处理方法

Adaptive filter processing method based on aerial images and terrain

点云滤波是点云数据处理中至关重要的一环,其结果往往影响到最终的点云数据质量。受植被及地形地貌的影响,点云滤波分类不准确,从而引起点云数据与真实的地形不一致。本文通过将点云数据与航空影像进行高精度配准后,内插出点云数据的RGB值,结合反射强度信息和高程信息进行植被滤除,将植被滤除后高低异常点的高程归算到拟合的二次曲面内进行处理。通过选取表面曲率、坡度、高斯曲率、平均曲率以及高程标准差等多个参数来描述地形的特征,通过地形特征参数进行地形自适应阈值滤波,并对以上研究方法采用了实测数据进行验证。结果表明:采用以上方法能对植被进行有效的滤除,点云滤波的精度有明显提高,采用该方法处理后的点云更接近真实地形。

Point cloud filtering is a crucial part of processing point cloud data,and its result often affects the final point cloud data quality.Due to the influence of vegetation and topography,the classification of point cloud filtering is often inaccurate,which causes the inconsistency between point cloud data and real terrain.In this paper,after high-precision registration of point cloud data and aerial images,the red,green,and blue(RGB)values of point cloud data were interpolated,and vegetation filtering was carried out by combining reflection intensity information and elevation information.The elevation of high and low anomaly points after vegetation filtering was calculated and fused into the fitted quadric surface for processing.Terrain features were described by selecting surface curvature,slope,Gaussian curvature,average curvature,and standard difference of elevation,and terrain adaptive threshold filtering was carried out according to terrain feature parameters.The results show that the vegetation can be effectively filtered by the above method,and the accuracy of point cloud filtering is obviously improved.The point cloud processed by this method is closer to the real terrain.