This paper focused on an analysis of the literature devoted to national and foreign land cover classification systems, which are based on different principles of building and using a different number of levels. A classification of land cover protected areas is based on the geobotanical research scientists of Saint Petersburg State University. Classification includes four levels. The proposed land cover classification of Saint Petersburg protected areas has been formulated on the basis of interpretation of remote sensing data. The land cover classification criteria are developed for its practical use.
This paper focused on remote sensing, digital acquisition and digital processing methods for the purpose of land cover automatic classification of Yuntolovsky reserve. It presents the technique of automatic identification of land cover according to different methods. Comparisons the results of maximum likelihood method and neural networks approach are presented in the paper.