深度学习在林业中的应用

林业数据信息的智能处理与深度挖掘分析是智慧林业精准决策与统筹管理的基础。文中主要从树木识别与分类、森林火灾识别与预测、树木病虫害监测和木材检测等4个方面总结深度学习在林业中的应用；从林业数据集、数据标记、特征学习涵盖范围、图像遮挡等方面分析了深度学习在林业领域应用上的缺点与局限性，以充分认识深度学习在林业应用方面亟待解决的问题与突破的方向；从林业领域应用场景、算法改进、数据集共享与研究成果应用的角度，展望深度学习在林业领域应用上亟待加强研究的方面，以提升深度学习解决林业应用问题的广度与深度，促进智能林业的发展。

Application of Deep Learning to Forestry

The intelligent processing and deep mining analysis of forestry data information are the basis for intelligent and precise forestry decision-makings and integrated management. This paper summarizes the application of deep learning to forestry from four aspects of tree identification and classification, forest fire identification and prediction, tree pest and disease monitoring and wood detection, and analyzes the shortcomings and limitations of deep learning application in forestry in terms of forestry datasets, data labeling, feature learning coverage, and image occlusion, so as to fully understand the problems and breakthrough directions of deep learning application in forestry. From the perspectives of application scenarios in forestry, algorithm improvement, dataset sharing and research results application, the future researches on the application of deep learning to forestry that need to be strengthened are prospected, in order to enhance the breadth and depth to deep learning application in forestry and promote the development of intelligent forestry.