基于MATLAB的DEM粗差探测算法比较分析

Comparison of DEM Gross Error Detection Algorithm Based on MATLAB

针对生成DEM时高程点的粗差探测问题,详细分析了Felicisimo加权平均算法、Felicisimo十字交叉算法、基于趋势面和三维可视化的粗差探测算法、基于坡度信息的粗差探测算法、基于点方式的粗差探测方法、抗差移动三次曲面拟合算法等常用的粗差探测算法的基本原理。在此基础上,利用MATLAB软件进行算法编程,通过对摄影测量建模所得的DEM格网数据中添加粗差的方式进行算法试验。结果表明:(1)在平原、山地等地势类似于平面或斜面的地形粗差探测中,Felicisimo十字交叉算法、Felicisimo加权平均算法、基于点方式的粗差探测算法和基于坡度信息的粗差探测算法具有较大优势;(2)在鞍部、山包、丘陵等地势接近于曲面的地形中,基于趋势面和三维可视化的粗差探测算法以及抗差移动三次曲面拟合算法具有较大优势。

Aming at the probing question of gross error elevation points existed in the establish process of digital elevation model( DEM),the technical principals of Felicisimo weighted average algorithm,Felicisimo crossing algorithm,gross detection algorithm based on trend surface and 3 D visualization,gross detection algorithm based on gradient information,gross detection algorithm based on point method and anti-differential movement cubic surface fitting algorithm are discussed in detail. Based on the above analysis results,the operation programme of the above algorithm is designed by MATLAB language,the gross error is added to the DEM grid data obtained by photogrammetric modeling,so as to analyze the algorithm performance. The results indicated that:(1) Felicisimo weighted average algorithm,Felicisimo crossing algorithm,gross detection algorithm based on trend surface and 3 D visualization,gross error detection algorithm based on gradient information are suitable for the gross error detection in plains,mountains;(2)the advantanges of other two algorithms are significant for detecting the gross errors in saddle,knoll and hills.