摘要：The paper presents a hybrid classification method based on the determination of the optimal number of classes according to uncontrolled classification followed by image processing techniques of controlled classification. A criterion for determining the optimal number of classes is proposed based on the definition of averaged values differences of average spectral brightness among the classes. Space images from satellites Ikonos (2002, 2007) and QuickBird (2010) were used to study different time cuttings in the forests of the Carpathian region. A significant amount of ground observation was held for getting test information. A Hybrid Classification Method is used for different time cuttings by QuickBird satellite images and implemented in a software environment of ERDAS Imagine. In order to obtain acreage of cuttings made for the period of 2002–2007 and 2007–2010, a comparative analysis of cuttings is introduced in these time intervals and their area is determined on the basis of the digital images of polygons in the ArcGIS software environment.
关键词：remote sensing of the Earth;space image;forest monitoring;classification;method of maximum probability;Mahalanobis’s method;method of minimum distance;deforestation
