Abstract: The study takes six urban districts as the research area. According to the forest land and forest stand classification standards in Detailed Regulations on Field Work of Forest Planning and Inventory in Shanxi Province, expert classfication method is adopted with the NFI data and SPOT5 remote sensing data, and 15 tree species and 16 shrubs and their combinations are identified successfully after the repeated tests. Without changing the original subcompartment division, the visual interpretation method is used to test and check the accuracy of the forest resources inventory data. The result shows that expert classification method can eliminate the influence of personal behavior on interpretation, which is better than visual interpretation in terms of details of interpretation. Expert classification method is more scientific and practical and has greater extention significance in aspect of production.

Key words: expert classification, remote sensing, forest resource, images of SPOT 5