使用高分遥感立体影像提取黄土丘陵区切沟参数的精度分析

发布时间：2018-07-01 09:22

本文选题：遥感 + 测量　；参考：《农业工程学报》2017年18期

【摘要】：为了研究高分立体像对测量黄土丘陵区切沟参数的适用性,选取陕北黄土区合沟小流域,以三维激光扫描全站仪获取的数据为参照值,分析使用Geo Eye-1高分遥感立体像对测量切沟参数的精度,得到如下研究结果。1)切沟面积、周长、沟长和沟宽等线状和面状参数平均测量误差分别为3.58 m~2,0.55 m,0.13 m和-0.10 m,其中面积、周长和沟长的百分误差主要集中在5%以内,沟宽百分误差主要分布在10%以内。2)切沟三维参数沟底宽、最大沟深、平均沟深的平均测量误差分别为-0.67、0.14和-0.46 m。截面积和体积的平均误差分别为-6.30 m~2和-54.01 m~3。最大沟深的百分误差主要集中在30%以内,沟底宽、平均沟深、截面积和体积的百分误差则主要分布在50%以内;相较于三维激光扫描的切沟,立体像对提取的切沟沟底形态误差较大,主要是沟底宽和平均沟深偏小。3)切沟规模越大,切沟体积、截面积和沟底宽的测量值偏小的幅度越大。但是,切沟体积测量误差与切沟体积之间可以建立较好的线性回归模型,在缺少其他测量手段时,可以使用该模型对测量误差进行校正。总体上看,高分立体遥感为切沟线状和面状参数测量以及切沟体积测量提供了新的方法,为黄土丘陵区沟蚀监测提供了便捷、且相对可靠的数据源。
[Abstract]:In order to study the applicability of high-score stereoscopic image in measuring the parameters of gully cutting in loess hilly region, the small watershed of Hegou in loess region of northern Shaanxi was selected, and the data obtained by 3D laser scanning total station was taken as reference value. This paper analyzes the accuracy of geo Eye-1 high remote sensing stereo image in measuring the parameters of cutting furrow, and obtains the following results. 1) the average measuring errors of linear and surface parameters such as area, circumference, length and width of trench are 3.58 mm2, 0.55 mm2, 0.13 m and -0.10 m, respectively, and the area of them is -0.10 m.The average measurement error is 3.58 mm2, 0.55 mt, 0.13 m and -0.10 m, respectively. The percentage error of circumference and furrow length is mainly within 5%, and the error of furrow width is mainly within 10%. 2) the mean measurement errors of furrow bottom width, maximum furrow depth and average furrow depth are -0.67m 0.14 and -0.46m respectively. The average errors of cross-sectional area and volume are -6.30 mm2 and -54.01 mm3, respectively. The percent error of the maximum furrow depth is mainly within 30%, the width of the furrow bottom, the average furrow depth, the percent error of cross section area and volume are mainly less than 50%. The larger the size of the furrow, the larger the volume, the smaller the measurement value of cross-sectional area and the wider the width of the furrow, the larger the shape error of the furrow bottom extracted from the stereo image is, mainly the width of the furrow bottom and the average furrow depth are smaller than that of the average furrow depth. However, a good linear regression model can be established between the error of cutting groove volume and the volume of cutting ditch, and the model can be used to correct the measurement error when other measurement methods are lacking. On the whole, high sublevel remote sensing provides a new method for the measurement of linear and surface parameters and volume of cutting trenches, and provides a convenient and relatively reliable data source for gully erosion monitoring in loess hilly areas.
【作者单位】：北京林业大学水土保持学院水土保持国家林业局重点实验室;
【基金】：国家自然科学基金项目(41671272) 国家重点研发计划(2016YFC0501604-05)
【分类号】：S127;S157.1

【相似文献】