【作者】 吴伶；

【导师】 刘湘南；

【作者基本信息】 中国地质大学（北京） ， 地图制图学与地理信息工程， 2010， 硕士

【摘要】 在农田生态系统中,铅是一种具有潜在危害的重要污染物,它不但会影响水稻的产量和品质,而且还可通过食物链富集等方式危害人类的健康。因此,研究水稻铅污染胁迫状况,分析铅污染水稻光谱曲线的遥感弱信息变化,对于防止粮食污染、保障人体健康有着重要的意义。本文重点研究基于分形-小波模型的水稻铅污染胁迫遥感弱信息增强与提取方法。在对分形理论、小波理论的分析、比较和总结的基础上,利用小波变换法计算分形维数实现对水稻铅污染胁迫遥感弱信息的提取。首先对健康水稻与铅污染胁迫水稻进行了奇异性分析,然后分别对健康水稻和铅污染水稻在分蘖期、抽穗期的原始高光谱曲线(450nm～700nm)及其二阶导曲线计算分形维数。结果表明,铅污染水稻的奇异性较健康水稻有所降低,所以铅污染水稻的原始高光谱曲线分形维数有所降低,同时铅污染水稻二阶导曲线分形维数较健康水稻也随之降低,且铅污染水稻二阶导曲线在分蘖期与抽穗期分形维数的差值比健康水稻的差值大,认为利用二阶导曲线的分形维数来对水稻是否受铅污染进行检测的效果比用原始光谱曲线的分形维数效果要好。同时还得出分蘖期是监测受铅污染水稻的最佳时期的结论。最后建立了基于分形维数的叶绿素浓度预测模型,并得到了不错的拟合效果。全文共分为五章。第一章阐明了利用遥感技术监测水稻铅污染胁迫水平的意义,在综述国内外研究现状的基础上,提出了本文研究的重点问题和解决方法。第二章总结了水稻铅污染的生化响应及遥感信息机理,遥感植被指数及其在农作物污染胁迫中的遥感应用,并指出这些遥感植被指数的不足。第三章介绍了实验样地、数据采集、研究波段的选择依据以及如何利用分形和小波理论来增强和提取受铅污染水稻的遥感弱信息。第四章对农田铅污染水稻光谱数据的弱信息进行增强和提取,分析增强的效果,并建立基于分形维数的叶绿素浓度预测模型,并根据实验结果分析该模型的性能优劣。第五章是对本文研究的总结,并在此基础上提出了今后进一步研究的设想。更多还原

【Abstract】 Pb is a potential hazardous pollution in farmland soil. Pb pollution has significant impacts on the growth of rice, and it may cause reduction of rice yield. The concentrations of pollutants will increasing as toxins are passed up the food chain and ultimately threatening the health of human beings. So studying of rice under Pb pollution stress and analysis of hyperspectral change of rice under Pb pollution stress are important for avoiding rice polluted and protecting agricultural environment and human being heath.The goal of this research is to derive a new model named fractal- wavelet model for extraction from remotely sensed information of rice under Pb pollution stress. According to analysis and compare and summarize of wavelet and fractal theory, this paper proposes the method of extracting hyperspectral weak change of rice under Pb pollution stress by using fractal dimension, which was performed by wavelet transform. First, singularity analysis were carried out between healthy rice and rice under Pb pollution respectively, and then for healthy rice and rice under Pb pollution respectively, the fractal dimensions of hyperspectral curves (450nm-700nm) and second derivative curves were calculated at the tillering stage and the heading stage. Results showed that the fractal dimensions of hyperspectral curves of the rice under Pb pollution decrease, compared with these of the healthy rice. The fractal dimensions of second derivative curves also decreased, and the difference in the fractal dimension between tillering stage and heading stage of the rice under Pb pollution is larger than that of healthy rice, according to which we regard that the fractal dimensions of second derivative curves in judging whether the rice was polluted is better. Meanwhile, it can be concluded that it is the best time to monitor the rice under Pb pollution at the tillering stage. Finally, the paper establishes a predict model for chlorophyll concentration based on fractal dimension, which we achieve a nice accuracy.There are five chapters in this paper. Chapter one is an overview of previous research work on the usage of remote sensing technology in vegetation health monitoring area. Chapter two introduces the biochemical reaction mechanism of rice under Pb pollution stress, and the vegetation indexes which are used in this field. Then it points the disadvantages of these indexes. Chapter three introduces the sample plot and data collection and the choice of range of band, and then how to enhance and extract spectral weak information with fractal and wavelet. Chapter four specifies the result of enhancing and extracting spectral weak information with fractal and wavelet, and then analysis the effect of the model. This paper establish a predict model for chlorophyll concentration based on fractal dimension at last. Chapter five is the conclusion of this research work and the plan of next work.更多还原