利用土壤表面灰度图像估算降雨过程中坡面土壤侵蚀量  

Using Soil Surface Gray Level Images to Estimate Soil Sediment Yields on Slope During Rainfall

作　　者：朱元骏[1] 倪盼盼 邵明安[1,3] 

机构地区：[1]西北农林科技大学黄土高原土壤侵蚀与旱地农业国家重点实验室,陕西杨凌712100 [2]中国科学院水利部水土保持研究所,陕西杨凌712100 [3]中国科学院地理科学与资源研究所,北京100101

出　　处：《水土保持学报》2014年第6期66-70,共5页Journal of Soil and Water Conservation

基　　金：国家自然科学基金项目(41371242);中国科学院"西部之光"项目;科技惠民计划项目(2012GS610203)

摘　　要：在室内模拟降雨条件下,利用数码相机获取土壤侵蚀过程影像,通过提取土壤表面图像建立图像灰度深度与侵蚀深度之间的关系来估算坡面土壤侵蚀量。结果表明：估算的土壤侵蚀量总体上要高于实测土壤侵蚀量,估算的相对误差在1.9%～38%之间,平均相对误差为18.9%,估算方法的精度为2.5%。该方法估算结果的归一化均方根偏差（Normalized root-mean-square deviation,NRMSD）为9.3%,表明其可以以较高的精度来估算降雨过程中的土壤侵蚀量。Constructing a method to estimate soil sediment yield quickly during rainfall could help to understand soil erosion process and pattern as well as to assess soil erosion status.Soil erosion process was videoed by a digital camera under laboratory rainfall simulation.Soil surface images were extracted from the video and a method that based on the relation between gray level range of the image and erosion depth was established to estimate soil sediment yield.The comparison of the estimated and the measured soil sediment yields reported that：The method generally overestimated soil sediment yield.Relative error of the method was 1.9%～38%,the average error was 18.9%,and the precision of the method was around 2.5%.Normalized root-mean-square deviation（NRMSD）for this method was 9.3%,indicated the method could estimate soil sediment yield during rainfall with a relatively high accuracy.

关 键 词：模拟降雨 灰度图像 图像分析 土壤侵蚀 

分 类 号：S157.1[农业科学—土壤学]