论文部分内容阅读
筏式养殖裙带菜监测是大连地区遥感监测的重要应用之一,寻求一种快速准确的识别方法从而及时掌握裙带菜养殖的分布范围、面积、数量等基础信息,对地方政府调控产业发展起到关键作用。引入数据挖掘技术,利用Landsat TM数据源,开展基于关联规则的裙带菜筏式养殖信息提取方法研究,并在大连金石滩附近海域进行实验论证。应用数据挖掘关联规则分析方法,找出筏式养殖区和海水分类的知识规则,并通过该规则构建分类树提取养殖信息,最后结合筏式养殖形态特征对得到的分类结果进行噪声去除后,得到最终的分类结果。结果表明:该方法的总体识别准确度可达80%,与最大似然分类结果相比识别准确度提高11.64%,该方法能够满足监测基本需求,具备一定的可行性。
Raft culture Undaria pinnatifida is one of the most important applications of remote sensing monitoring in Dalian. It is necessary to find a fast and accurate method to identify the distribution, area and quantity of Undaria pinnatifida and other basic information, and to control the development of local government Key role. With the introduction of data mining technology and Landsat TM data source, this paper studied the extraction method of raft culture information based on association rules, and carried out experimental demonstration in the sea area near Jinshitan, Dalian. Using data mining association rules analysis method, the knowledge rules of raft culture area and seawater classification are found out, and the classification tree is constructed by this rule to extract the farming information. Finally, after combining the raft culture morphological characteristics to remove the noise of the classification results, The final classification result. The results show that the overall recognition accuracy of the proposed method is up to 80%, and the recognition accuracy is improved by 11.64% compared with the maximum likelihood classification. This method can meet the basic monitoring needs and has certain feasibility.