Mining is an extractive activity of great socioeconomic importance, as mineral products are a promising source of wealth and employment for many families. However, as a result of human activities, mining causes significant ecosystem degradation, posing several challenges for rehabilitation.This study aims to monitor and map the progress of pastureland rehabilitation into secondary forest through vegetation compensation, using object-based image analysis (GEOBIA) techniques applied to high-resolution satellite imagery. The study focuses on rehabilitation blocks 1 to 6 and the São Luiz farm, located in Canaã dos Carajás and Ourilândia do Norte, in the southeastern region of Pará state, Brazil. This approach allows for the analysis of vegetation and soil spectral behavior, as well as the calculation of spectral indices such as NDVI. The methodology applied to the rehabilitation areas contributes to improving environmental monitoring strategies and ecological rehabilitation.The Salobo project, monitored by Vale, carries out environmental compensation through the rehabilitation of degraded areas in Canaã dos Carajás and Ourilândia do Norte. The research utilized CBERS-4A and Sentinel-2A satellite images to monitor vegetation using object-based image analysis (GEOBIA) and NDVI calculations. The processing involved atmospheric correction, segmentation, and classification of images into four categories: exposed soil, grasses, shrubs, and forests. The data were processed in a Geographic Information System (GIS) for environmental recovery analysis.The study focused on mapping three vegetation classes along with exposed soil in areas undergoing environmental rehabilitation, which were previously used for livestock activities. Three time periods were analyzed—2017, 2020, and 2024—using CBERS-4 and Sentinel-2 images. Classification was based on the NDVI index to identify vegetation areas with different levels of vigor. Segmentation was performed at different scales to improve accuracy, particularly in areas with exposed soil and unhealthy vegetation (grasses).The results showed a significant increase in healthy vegetation over time and a reduction in exposed soil across all areas. In 2017, exposed soil in blocks 1, 2, 3, and 4 accounted for 84%, 92%, 82%, and 72%, respectively. By 2024, these areas had reduced to just 0.3%, 1.2%, 0.6%, and 2.4%, with a remarkable increase in forested areas to over 90% in all locations. The use of NDVI allowed for tracking the recovery process and evaluating the effectiveness of soil management practices, highlighting the importance of spectral indices in environmental monitoring. The results indicate that the studied areas have been revegetated and monitored via GEOBIA, reinforcing the importance of remote sensing for monitoring degraded areas.