Sovremennye problemy distantsionnogo zondirovaniya Zemli iz kosmosa, 2009, Issue 6, Vol. 2, pp. 373-379
Estimation of the possibilities of forest forming species recognition
on the base of IKONOS space images
В.М. Жирин
, С.В. Князева
Центр по проблемам экологии и продуктивности лесов РАН (ЦЭПЛ РАН), 117997 Москва, ул. Профсоюзная, 84/32
The results of experiment on the study of the main species recognition possibilities of IKONOS high resolution
images was carried out on the example of the National Park Losinyi ostrov are represented. Supervised
classification by the method of maximum likelihood makes possible to recognize the generalized groups of deciduous
and coniferous stands with the accuracy of more than 80% and to calculate the composition coefficients
of the deciduous and coniferous stands in the range of the taxation site with accuracy to 2 units. The application
of derived images makes it possible to increase the recognition certainty of the separate tree species, but does not
improve deciphering the generalized groups of deciduous and coniferous stands.
Keywords: the forest forming species, the stand composition coefficients of the taxation site, coniferous and deciduous stands, the high spatial resolution space images, classification by the method of maximum likelihood, textural characteristics, derived images, the recognition certainty of classes
Full text