结合属性邻接图与点云的零件模型特征识别方法
A part model feature recognition method combining attribute adjacency graph and point cloud
舒敏 1杨涛2
作者信息
- 1. 西南科技大学信息工程学院,绵阳 621010
- 2. 特殊环境机器人技术四川重点实验室,绵阳 621010
- 折叠
摘要
针对目前基于属性邻接图与点云的零件模型特征识别技术存在的局限性,结合2种特征识别方法提出了一种结合属性邻接图与点云的零件模型特征识别方法.利用模型属性邻接图匹配特征子图找到特征面并分离,再将特征面进行点云采样,最后在PointNet网络基础上改进点云分类网络结构.通过添加局部特征提取模块与基于Transformer网络的非局部特征提取模块,并结合特征属性邻接图信息与原始点云数据,对24种常见特征进行特征识别试验,最终识别准确率为 99.92％.
Abstract
A part model feature recognition method combining attribute adjacency graph and point cloud was proposed by combi-ning two feature recognition methods to overcome the limitations of current part model feature recognition technology based on at-tribute adjacency graph and point cloud.The model attribute adjacency graph was used to match feature subgraphs to find and sep-arate feature surfaces,and then the feature surfaces in point clouds were sampled.The point cloud classification network structure on the basis of PointNet network was improved by adding a local feature extraction module and a Transformer based non-local fea-ture extraction module and combining feature attribute adjacency graph information with original point cloud data.Experimental re-sults indicate that the recognition accuracy for 24 common features is 99.92％.
关键词
零件模型特征识别/属性邻接图/点云/Transformer网络/PointNetKey words
part model feature recognition/attribute adjacency graph/point cloud/Transformer net/PointNet引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据