In the conflict-affected countries, increasing realization of the fact that forests not only provide multiple benefits to local communities but also help in conserving the environment and mitigating climate change has created global concern for their protection and conservation. However, if these resources have not been managed in a sustainable or equitable manner, this leads to further environmental degradation and global warming. In this direction, vegetation mapping may be a primary requirement for various management and planning activities at the landscape level. The study presented here focused on developing methods of tree species identification in conflict areas using aerial hyperspectral data. Five RapidEye scenes were acquired for that purpose in the Nuba Mountains of Sudan. The Geographic Object-Based Classification (GEOBIA) i.e., K-Nearest Neighbor classifier model and knowledge-based classifier, built on a developed model of integrated features (such as vegetation indices, DEM, and thematic layers), environmental knowledge, feature extracted from RapidEye images, and user expert knowledge, was applied to generate the species map. The overall accuracy, producer’s accuracy, user’s accuracy, and Kappa statistics methods were conducted. Additionally, for more accurate results of each class, the best classification result method was also applied. GEOBIA provides unprecedented opportunities to classify and detect tree species more accurately, over large areas, with diminishing costs and processing time. The study recommends that the use of GEOBIA to analyze RapidEye imagery for the identification of trees species in the semi-arid region should be integrated with ancillary data such as DEM and other levels of GIS data to improve the quality of the results.