王晓红,谌鹏,刘芳,杜景敏,许禾昕,唐丰圆,方加炜.基于集成学习的强鲁棒性三维点云数据分类研究[J].包装工程,2021,42(3):252-258.
WANG Xiao-hong,CHEN Peng,LIU Fang,DU Jing-min,XU He-xin,TANG Feng-yuan,FANG Jia-wei.Robust 3D Point Cloud Data Classification Based on Ensemble Learning[J].Packaging Engineering,2021,42(3):252-258. |
| 基于集成学习的强鲁棒性三维点云数据分类研究 |
| Robust 3D Point Cloud Data Classification Based on Ensemble Learning |
| 投稿时间：2020-09-02 |
| DOI：10.19554/j.cnki.1001-3563.2021.03.036 |
| 中文关键词: 检测 识别 包装 生产 点云分类 集成学习 中心损失 鲁棒性 |
| 英文关键词: detection identification packaging production point cloud classification ensemble learning central loss robust |
| 基金项目:上海市科学技术委员会上海市自然科学基金面上项目(19ZR1435900);“基于柔印产品特性的智能化印前图像处理”招标课题(ZBKT201809) |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 通过三维扫描仪得到的点云数据往往存在很多异常值，例如噪点、遗失点和外部点等。在这些异常值存在的情况下，为了提高三维点云数据的分类精度，提出一种基于集成学习的强鲁棒性三维点云数据分类方法。方法 提出一种基于最大投票法的集成学习思想，将2个深度神经网络的分类结果进行集成，从而提高网络的泛化性和准确性;采用全局特征增强和中心损失函数来优化神经网络结构，提高分类精度并增强鲁棒性。结果 文中方法缩短模型训练时间至30个迭代次数，且在有噪点、丢失点和外部点的情况下分类精度均得到有效提升。结论 提出的EL-3D算法在含有噪点、丢失点和外部点的情况下，鲁棒性效果要优于目前的点云分类方法。 |
| 英文摘要: |
| There are many outliers in the point cloud data obtained by 3D scanner, such as noise, missing points and external points. The work aims to propose a strong robust 3D point cloud data classification method based on ensemble learning to improve the classification accuracy of 3D point cloud data in the presence of these outliers. An idea of ensemble learning based on maximum voting method was proposed, which integrated the classification results of two deep neural networks to improve the generalization and accuracy of the network; global feature enhancement and central loss function were used to optimize the neural network structure to improve the classification accuracy and robustness. The proposed method shortened the training time to 30 epochs, and improved the classification accuracy effectively in the case of noise, missing points and external points. The proposed el-3d algorithm has better robustness than the current point cloud classification methods in the case of noise, missing points and external points. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
渝公网安备 50010702501716号