Using images combined with DEM in classifying forest vegetations

More and more different resolution images are used in the process of forests vegetations classifying and recognizing. Most classifying methods, including supervising and non-supervising methods, are based on spectral information and work well in forest vegetations recognizing. But some forest vegetations such as conifer, broadleaf and mixed forest are not distinguished clearly by using these methods just because these vegetations have similar spectral characteristic. In this paper, TM image of Changbai Mountains was obtained and spectral information of the image is analyzed. At the same time, the vegetation types covering the Changbai Mountains are acquired by investigating the region in person. On the other hand, DEM (digital elevation model) is employed as an important criterion in classifying forest vegetations basing on spectral character in this article. The classified results are better than those calculated by original classifying methods. At last, the results are verified by data surveyed on the spot. The results illustrate that DEM is an important factor in classifying forest vegetations distributed by the elevation. Spectrum combined with elevation information is very useful for object recognizing and classifying. Images and DEM have widely practical application and should be used in more research fields.