扎龙湿地沼泽的提取模型研究

MARSH INFORMATION EXTRACTION BASED ON KNOWLEDGE DISCOVERY

利用扎龙湿地TM图像,通过分析各类地物的光谱特征、谱间结构特征以及采样剖面线,发现了扎龙湿地沼泽的光谱特征,并由此建立了一个基于知识发现的沼泽提取模型。提取结果表明,建立的沼泽提取模型能够实现自动提取扎龙湿地的芦苇沼泽,其精度可以达到98%。

This paper deals with the way of extracting reed marsh information in Zhalong wetland based on the knowledge of spectrum characteristics. Spectral analyses of TM image covering Zhalong wetland were conducted, which included analysis of spectrum average value characteristics, analysis of characteristics of different spectrum structures and analysis of the sampling profile. This led to the discovery of the spectrum character knowledge of the marsh in Zhalong wetland. With the results obtained, a model of marsh extraction based on knowledge discovery was established. The marsh information extraction shows that the model proposed in this paper can extract reed marsh in Zhalong wetland automatically, and the extraction accuracy can reach 98%.