基于决策规则的遥感影像土地利用信息提取

遥感作为近几十年迅速发展起来的一门综合性技术学科,己经在许多领域发挥了重大作用。通过对遥感数据进行专题分类处理以得到土地利用等专题信息是遥感最广泛的应用领域之一。尽管土地利用遥感分类方法不断发展,但分类技术始终跟不上遥感技术本身的发展。本论文的主要目标之一就是以沈阳矿区为研究区,利用多源遥感数据结合地面实地调查,围绕遥感图像,采用常规的最大似然分类法,同时采用选用决策树分类方法,对不同数据源得到的信息进行综合分析,充分利用其中的光谱信息、地学知识以及人的经验知识进行土地利用分类,从而更好地为地方土地有效利用提供决策依据。

The land utilization information extraction of remote sensing image based on decision rule

Remote sensing,as a comprehensive technological disciplines of developing rapidly near decades,has been played a great role in a lot of fields.Through carries on special classification processing to the remote sensing data to obtain special information of land utilization and so on is one of the most extensive applications of remote sensing.Though the land utilizes taxonomic method of remote sensing develops constantly,taxonomic technology can not catch up with the development of remote sensing.One of the main targets of this paper is that utilizing the remote sensing data of many sources to combine the ground field investigation with the mining area of Shen yang for the studying area,adopt the regular maximum likelihood classifier,at the same time adopt the decision-tree clas- sification,to carry on comprehensive analysis to the information received from the different data source,fully spectrum information which utilize among them,gain knowledge and utilize classify land at experience knowledge of people.The full using of spectrum infor- mation,the geoscience knowledge as well as person's experience knowledge carries on the land utilization classification,thus providing the policy-making basis for the place effective land use.