Unmanned Aerial Vehicles (UAVs) play a crucial role in exploring unpredictable terrains, such as accident sites and search zones. These robots require autonomous navigation and precise path planning to operate safely and efficiently in dynamic environments. UAVs must balance navigation accuracy and energy efficiency constraints during autonomous operations. This paper introduces an innovative boundary-driven mapping strategy for UAV exploration in unknown environments. The study proposes a novel approach for boundary extraction using deep learning and presents a decision-making methodology to enable the UAV to select the most optimal frontier for exploration based on deep learning-derived information. The primary objective is to efficiently expand the unknown map. The research demonstrates a significant reduction in exploration time within a simulated environment, achieving better performance compared to other methodologies documented in the literature. The results highlight the effectiveness and efficiency of the proposed strategy in terms of mapping exploration.