出版社：Grupo de Métodos Cuantitativos, SIG y Teledetección, A.G.E
摘要：This paper analyzes the use of CBERS (China-Brazil Earth Resources Satellite) high resolution images to map urban limits of a small city in the Amazon Region of Brazil. The use of Sino-Brazilian satellite images to study small cities of the Brazilian Amazon Region is also discussed. A HCR image was used to classify the image of the study area into four classes: urban, forest, clouds, and shadows, by visual interpretation. This map was compared with another map produced with unsupervised classification. A confusion matrix was produced to evaluate the accuracy of unsupervised classification. Comparison between the reference map (visual interpretation) and the unsupervised classification map verified that 82.8% of the areas were classified correctly. This ability to recognize and delimitate small cities in Amazonia demonstrates the quality of the CBERS – HCR image.
关键词：CBERS- HCR image, high resolution image, urban area delimitation, Brazilian Amazon towns.
