Forest fires cause catastrophic impacts on ecosystems, economies, and human health, necessitating advanced detection methods. Climate change and erratic weather patterns exacerbate these infernos, which consume vast land areas, destroy wildlife, and release toxic pollutants into the atmosphere. Traditional methods relying on delayed satellite imagery or human observation fall short in providing timely alerts. This study proposes a novel system integrating IoT sensor networks and machine learning to address these challenges. A regression model trained on real-time IoT sensor data predicts potential fire events, while a YOLO-based object detection model visually confirms fire occurrences. This dual-confirmation approach enhances detection accuracy and reduces false positives. Furthermore, the incorporation of unmanned aerial vehicles (UAVs) guarantees timely reactions and facilitates effective monitoring of expansive forest regions. Experimental results demonstrate the system’s effectiveness in early and reliable fire detection, paving the way for improved fire management strategies and the protection of natural resources.