Land-cover classification of an intra-urban environment using high-resolution images and geographic object-based image analysis : the case of APA Mananciais do Rio Paraíba do Sul

Protected areas of sustainable use such as the Environmental Protection Areas (APA) encompass urban areas. Because the characteristic urban spaces are under dynamic changes, they usually entail problems related to planning land cover. Such areas are fragile, especially when located inside protected areas, so it is necessary to monitor and evaluate them. Remote sensing data provides important information for urban planning and management issues, and have a great potential to assist conservation unit managers in monitoring such protected areas. Urban environments are characterized by high spectral and spatial heterogeneity and, consequently, most urban pixels in moderate resolution imagery contain multiple land-cover materials. The objective of this paper is to demonstrate the capability of RapidEye sensor data, for the intra-urban scale classification of land cover in protected areas, and to develop a semi-automatic classification method based on geographic object-based image analysis and data mining techniques, for efficiently identifying small changes in urban areas . The APA of “Mananciais do Rio Paraíba do Sul” (APA-MRPS), aimed to preserve the water sources for more than 15 million people, was selected as study site. The results showed that RapidEye data and the methodology used were effective in classifying constructed areas, enabling the identification of small changes in land cover. The data and methodology may be able to assist managers in the monitoring and evaluation processes of protected areas, especially APAs.