点群データの特徴量抽出

• 手島 裕詞

  佐世保工業高等専門学校

• 志久 修

  佐世保工業高等専門学校

• 金谷 孝之

  広島国際大学医療経営学部

• 西尾 孝治

  大阪工業大学情報科学部

• 小堀 研一

  大阪工業大学情報科学部

書誌事項

タイトル別名

• A Method for Extracting Features from Point Clouds

抄録

Point clouds have been used as a new data structure for 3D models in the fields of 3D computer graphics and computer-aided design. In this study, we propose a method for extracting the features from a point cloud based on the salience of a shape (as a basic research). The proposed method can be applied for shape analysis, recognition, similarity retrieval, and so on. This method extracts the features from a point cloud without constructing a polygonal mesh. First, the method calculates a variance vector using the differences between the normal vectors of each point. Next, using the distance and variance values, the feature is computed as the Y-intercept of linear regression. Several experimental results show that the method is effective for extracting the features from point clouds.

収録刊行物

• 設計工学

  設計工学 51 (2), 107-117, 2016

  公益社団法人 日本設計工学会

キーワード

• point cloud
• CG
• CAD
• shape feature
• linear regression

詳細情報

• CRID: 1390282680733071872
• NII論文ID: 130005125339
• DOI: 10.14953/jjsde.2015.2642
• ISSN: 21889023
• ISSN: 09192948
• 本文言語コード: ja
• データソース種別:
  • JaLC
  • CiNii Articles
  • KAKEN
• 抄録ライセンスフラグ: 使用不可