Pine wilt disease spreading prevention system using semantic segmentation

Pine wilt disease is a that affects ecosystems by rapidly killing trees in short period of time due to the close interaction between three factors such as trees, mediates, and pathogens. There no 100% mortality infectious forest pests. According Korea Forest Service survey, April 2019, damage pine re-nematode was about 490,000 dead 117 cities, counties wards across country. It's fatal condition. In order prevent this problem, paper proposes system detects early infection like, using deep learning-based semantic segmentation. addition, drones were used photograph area forest, separate pixel segmentation label could be identify levels transmission information: Suspicion, attention, confirmation. This allows user grasp information area, location, alarm spread disease.