Grassland Classification Based On Multispectral Image Data

As the most important land type in Inner Mongolia,grassland is the absolute main body of Inner Mongolia ecological environment.Different types of grassland are determined by local climate,temperature,water source,soil,topography and other geographical conditions.Classification of grassland is conducive to the dynamic monitoring of local grassland by agriculture,forestry,transportation and other departments.It is not only conducive to distinguish land use,but also of great significance to the ecological construction and protection of Inner Mongolia.In this paper,Hailar grassland in Inner Mongolia is taken as the research area,and the grassland classification methods are compared.First,analyze the advantages and characteristics of multispectral image data,extract the vegetation index reflected by different bands,and retain the vegetation data to be classified;then,analyze the implementation method of ant colony algorithm applied to multispectral image data classification,and complete the classification by feature search.Finally,in order to prove the effectiveness and efficiency of using ant colony algorithm to achieve grassland classification of multispectral image data,this paper compares the nearest neighbor method and the traditional maximum likelihood method.The final experimental results show that this method has high accuracy and efficiency in grassland classification of multispectral image data.This paper has 16 figures,8 tables and 52 references.