Multimodal classification of forest biodiversity potential from 2D orthophotos and 3D airborne laser scanning point clouds
基于2D正射影像和3D空中激光扫描点云的森林生物多样性潜力多模态分类
机构 * Perception Laboratory, Aalborg University, Denmark ; Pioneer Centre for Artificial Intelligence, Denmark ; Department of Computer Science, Copenhagen University, Denmark ; Institute for Visual \& Analytic Computing, Rostock University, Germany ; University of Barcelona ; Computer Vision Center, Spain
专题命中 点云 :point cloud(title,abstract);分类 cs.CV
AI总结 本研究利用2D正射影像和3D ALS点云数据,通过多模态深度学习融合方法,实现对森林生物多样性潜力的高效评估,实验结果达到82%的准确率。