【摘要】 松材线虫病是由松材线虫（Bursaphelenchus xylophilus）引起的一种危害严重的森林病害，对森林生态系统造成严重威胁。该研究旨在应用高分二号卫星监测变色松树，并评估应用效果，构建基于深度学习的监测体系，以提高监测的时效性和精确度。研究结合卫星遥感影像数据和无人机航拍影像，在广东省广宁县开展监测试验，影像的处理结合GS融合和GNDVI指数，显著提升变色松树与背景间的差异。使用基于深度学习算法进行变色松树提取，得到精确率为94.65%，召回率为83.30%,F1分数为88.61%。结果表明，基于高分二号影像的识别方法的精度较高，能为松材线虫病的精准监测工作提供数据支撑。更多还原

【Abstract】 Pine wilt disease, caused by pine wood nematode(Bursaphelenchus xylophilus), is a devastating forest disease that poses severe threats to forest ecosystems. This study aims to apply GF-2 satellite imagery for monitoring discolored pine trees and evaluate its effectiveness by establishing a deep learning-based monitoring system to enhance timeliness and accuracy.Integrating GF-2 satellite remote sensing data with UAV aerial imagery, monitoring experiments were conducted in Guangning County, Guangdong Province. Image processing techniques, including Gram-Schmidt(GS) fusion and the Green Normalized Difference Vegetation Index(GNDVI), significantly improved the contrast between discolored pine trees and their background.Using a deep learning algorithm for discolored pine tree extraction, the method achieved a precision of 94.65%, recall of 83.30%,and F1-score of 88.61%. The results demonstrate that the GF-2-based identification method offers high accuracy and can provide reliable data support for precise monitoring of pine wilt disease.更多还原