This paper proposes an intelligent wildfire detection system that combines IoT sensor nodes with deep learning-based video analysis. The proposed system uses low-power sensor nodes equipped with temperature, humidity, smoke, and gas sensors, powered by solar energy. LoRa-based mesh communication enables data transmission to a central server for real-time analysis. Additionally, deep learning models (CNN and LSTM) are applied to environmental data to predict potential wildfires. Experimental validation confirms the system’s effectiveness in early fire detection, enabling rapid response and reducing potential damage.

• 정재종(인하공업전문대학 기계설계공학과 재학생) | Jae-jong Jeong (Student, Dept. of Mechanical Design Engineering, Inha Technical College) Corresponding author
• 하재현(인하공업전문대학 기계설계공학과 재학생) | Jae-hyun Ha (Student, Dept. of Mechanical Design Engineering, Inha Technical College) Corresponding author
• 이정익(인하공업전문대학 기계설계공학과 교수) | Jeong-ick Lee