作　　者： (赵伶俐）; (王福平）;

机构地区： 北方民族大学电气信息工程学院,银川750021

出　　处： 《节水灌溉》 2017年第8期17-20,共4页

摘　　要： 为探讨利用土壤图像灰度和蒙特卡洛方法结合测量土壤含水量的可行方法,以宁夏引黄灌区为研究区,采集土壤图像样本提取灰度数据,分析土壤图像灰度的概率分布,构建以土壤图像灰度值为输入土壤含水量为输出的数学模型,采用蒙特卡洛方法测量土壤含水量,对所得土壤含水量检测值与实测值进行精度验证对比。试验结果表明:(1)用蒙特卡洛方法测量土壤含水量,减少了复杂非线性模型的分析,减少计算一些附加的一阶或高阶灵敏度系数,降低计算量。(2)测量的精度取决于蒙特卡洛试验次数,当增加蒙特卡洛试验次数,土壤含水量测量值越精确。(3)对蒙特卡洛法所得测量值进行克里格插值制图分析土壤含水量的空间分布格局,从仿真结果看较为符合研究区土壤含水量的实际状况。 In order to explore the feasible ways to measure soil water content by using soil image gray level data and Monte Carlo method, Helan County of Ningxia Yellow River Irrigation Area is selected as the research area, soil picture is gathered and the gray probability of soil image is analyzed, then a mathematical model, with the soil moisture content as output and the soil image gray-values as input, is established, the soil moisture is calculated by Monte Carlo Method, and the calculation value is compared with the measured value. The experimental results show that ： （1） using Monte Carlo method to measure the soil moisture content reduces the analysis of complex nonlinear model, the calculation of some additional first or higher order sensitivity coefficients, and the computational complexity. （2） The accuracy of the measurement depends on the number of Monte Carlo tests. The more accurate the Monte Carlo test, the more accurate the soil moisture measurement. （3） The results of using Monte Carlo method to analyze the spatial distribution pattern of soil water content are accorded with the actual situation of soil water content in the study area.

关 键 词： 土壤含水量检测 图像灰度 蒙特卡洛方法