中国机械工程

• • 上一篇    下一篇

[材料成形中的数字化与智能化产品设计]法向算子和D2算子相结合的铸件三维模型分类算法

孙晓龙1;张志鹏1;计效园1;童家良2;管雅倩2;张焕东3;周建新1   

  1. 1.华中科技大学材料成形与模具技术国家重点实验室,武汉,430074
    2.安徽华峰重工机械有限公司,宣城,242053
    3.江苏万恒铸业有限公司,盐城,224500
  • 出版日期:2020-11-25 发布日期:2020-11-27
  • 基金资助:
    国家自然科学基金资助项目(51905188,51775205)

Classification Algorithm of Casting 3D Model Based on Normal Operator and D2 Operator

SUN Xiaolong1;ZHANG Zhipeng1;JI Xiaoyuan1;TONG Jialiang2;GUAN Yaqian2;ZHANG Huandong3;ZHOU Jianxin1   

  1. 1.State Key Laboratory of Materials Processing and Die & Mould Technology,Huazhong University of Science and Technology,Wuhan,430074
    2.Anhui Huafeng Heavy Industry Machinery Co., Ltd.,Xuancheng,Anhui,242053
    3.Jiangsu Wanheng Casting Co.,Ltd.,Yancheng,Jiangsu,224500
  • Online:2020-11-25 Published:2020-11-27

摘要: 为了解决铸造生产中相同类型铸件的历史工艺复用问题,提出了一种法向算子和D2算子相结合的铸件三维模型分类算法。该方法从三维模型的面特征和点特征信息出发,分别提取三维模型的法向算子和D2算子特征描述符,通过度量新铸件模型和各类别模型特征描述符的相似性实现铸件模型分类。在企业实际生产的铸件模型库中的实验结果表明,该方法对常见的铸件类别有很好的特征识别效果,解决了D2算子作为预分类器时处理部分复杂模型时存在的特征不敏感的缺陷,可以获得准确率高的分类结果。

关键词: 三维模型分类, 法向算子, D2算子, 特征描述符, 铸件三维模型

Abstract: In order to solve the problems of historical process reuse of the same casting types in casting productions, an algorithm of casting 3D model classification combining normal operator and D2 operator was proposed herein. In this method, the normal operator and D2 operator feature descriptors of the 3D models were extracted from the surface features and point feature information of the 3D models. The 3D model classification was realized by measuring the similarity between the casting feature descriptors and the basic feature descriptors of each category. The experimental results of the casting model library actually produced by companies show that the classification algorithm combining the normal operator and the D2 operator has a good feature recognition effectiveness for the common categories of castings, and solves the defects of feature insensitivity when D2 operator is used as a pre-classifier to deal with part of complex models, and may obtain classification results with higher accuracy.

Key words: 3D model classification, normal operator, D2 operator, feature descriptor, 3D model of casting

中图分类号: