摘要土壤盐渍化是最常见的土地退化过程,目前已成为了全球性问题,利用土壤的电磁波谱信息,遥感技术能够获取广域、多波段、多时相的土壤信息,使大面积实时动态监测盐碱化状况成为可能。本文对国内外土壤盐渍化研究进展作了介绍,对基于高光谱的土壤盐分反演研究方法进行了总结。52870
以新疆维玛纳斯县作为研究区,在获取盐碱土高光谱数据的基础上,着重研究了盐渍土盐分含量光谱特征变化规律;对光谱进行微分、倒数对数及对数指数等提取了敏感波段,建立了光谱特征参数;通过二阶最小二乘法和曲线拟合等方法,对盐渍土盐分进行模型反演,并经过对比分析和精度验证,选定了最优模型。经过研究发现,土壤盐分的敏感波段为:998nm-1000nm、1399nm -1404nm、1890nm-1895nm、2200nm -2205nm。不同的光谱数据处理方法所获取的敏感波段略有差异。在此基础上,对土壤盐分含量进行了基于高光谱的建模,取得了较好的反演效果。
毕业论文关键词:土壤盐渍化;高光谱遥感;盐渍土光谱特征;土壤盐分浓度
Abstract Soil salinization is one of the most common land degradation process. At present, it has become the global problem, using the soil of the electromagnetic spectrum information and remote sensing technology to obtain wide area, multi band, multi phase of soil information, make large area real-time dynamic monitoring saline conditions possible. In this paper, the research progress of soil salinization at home and abroad is introduced, and the method of inversion of soil salinity based on hyperspectral data is summarized in this paper.
To Xinjiang Uygur Manas County as the study area, on the basis of saline alkali soil hyperspectral data acquisition, focuses on the saline soil salt content of spectral features variation; using spectral derivative method, the logarithm of the reciprocal and correlation analysis extracted sensitive bands and spectral characteristic parameters is established; by second-order least square method and curve fitting method, on soil salinity content of inversion model, and through comparative analysis and accuracy verification, and selected the optimal model. It is found that the soil salinity sensitive bands: 998nm-1000nm, 1399nm, -1404nm 1890nm-1895nm, 2200nm -2205nm. Different spectral data processing method for the sensitive band is slightly different. On this basis, the soil salt content was modeled based on Hyperspectral inversion, has achieved good effect.
Keywords: soil salinization; hyperspectral remote sensing; spectrum characteristics of saline soil; soil salt concentration
目 录
第一章 绪论 6
1.1研究背景与意义 6
1.2土壤盐渍化国内外研究进展 7
1.2.1国外研究情况 7
1.2.2 国内研究情况 8
1.3 基于高光谱的土壤盐分反演方法介绍 9
1.3.1偏最小二乘回归 9
1.3.2人工神经网络 10
1.3.3多元线性回归 10
1.3.4其他方法 11
1.4 研究内容和技术路线图 11
第二章 研究区概况 12
2.1研究区概况 13
2.1.1地理位置 13
2.1.2地形地貌 13
2.1.3水系 14
2.1.4气候