The use of Deep Learning tools for tree identification has been steadily increasing in the last 10 years, driven by advances in computational power and image analysis techniques. Deep neural network models, particularly architectures like CNNs (Convolutional Neural Networks), are effective for analyzing aerial, drone, and satellite images to recognize distinct canopy patterns and specific tree species. This project focuses on the identification of two native species of Brazilian flora found in the Cerrado biome: Buriti (Mauritia flexuosa) and Palmito Juçara (Euterpe edulis). To achieve this, were applied state-of-the-art Deep Learning techniques, specifically the RT-DETR and YOLOv8 methods, which excel in object detection tasks. A comprehensive data set was created consisting of high-resolution images captured by aerial drone surveys. Using this data set, the project’s goal was to demonstrate the potential of deep learning to automate the identification of plant species, ultimately contributing to ecological monitoring and conservation efforts in Brazil’s rapidly changing ecosystems.