Three-dimensional convolutional neural network model for tree species classification using airborne hyperspectral images
收藏国家林业和草原科学数据中心2022-11-16 更新2024-03-06 收录
下载链接：
https://www.forestdata.cn/dataDetail.html?id=CSTR:17575.11.0220221116246.040001.V1
下载链接
链接失效反馈官方服务：
资源简介：
该论文以高峰林场为研究区，针对机载高光谱数据的特点，通过将全连接层替换为一维卷积层构建了3D-1D-CNN模型，在保证精度的前提下减少训练时间，实现复杂地形下多树种的短时间大面积高精度分类，有很大的林业实际生产潜力。
This paper takes Gaofeng Forest Farm as the study area. Aiming at the characteristics of airborne hyperspectral data, a 3D-1D-CNN model was constructed by replacing fully connected layers with 1D convolutional layers. This model reduces training time while maintaining classification accuracy, enabling rapid, large-area and high-precision classification of multiple tree species in complex terrain, and has great practical production potential in forestry.
提供机构：
国家林业和草原科学数据中心
创建时间：
2022-11-16
