重回帰モデルによるブラジャー着装シミュレーションとその評価

• 崔 童殷

  京都工芸繊維大学大学院先端ファイブロ科学専攻

• 中村 顕輔

  京都工芸繊維大学大学院先端ファイブロ科学専攻

• 黒川 隆夫

  京都工芸繊維大学大学院先端ファイブロ科学専攻

書誌事項

タイトル別名

• Simulation of Brassiere-Wearing Figures Using Multi-Regression Models and Its Evaluation
• ジュウ カイキ モデル ニ ヨル ブラジャーチャクソウシミュレーション ト ソノ ヒョウカ

抄録

This paper proposes a method of simulating breast shape modification by wearing a brassiere. For this purpose a three-dimensional (3D) human body shape model developed by one of the authors is used. The model is made of a bi-cubic B-spline surface and its control points can describe a woman's trunk shape by fitting it to plenty of her body surface points. Corresponding control points among models have the same meaning in the sense that they form the same region on the different women's body surfaces. This feature enables us to compare and analyze a local body shape using a subset of control points selected appropriately. Forty-nine control points were determined to analyze breast shape of any woman. One hundred forty-two Japanese women aged 20's to 50's (brassiere size ranged from A70 to E70) were measured by means of an optical range finder before and after putting on a brassiere. We also categorized softness of their breasts into three classes; "soft", "medial", and "stiff". The multiple regression analysis established the relationship between the position of the control points on the models before and while wearing a brassiere using softness as a parameter. The best regression function was chosen among prepared 15 ones for each coordinate of every control points and was applied to simulation of brassiere-wearing breast figures based on the position of the control points before putting on it. The results showed that it is possible to estimate the body shape when wearing a brassiere. The same was confirmed based on average errors between silhouettes of brassiere-wearing and simulated breasts.

収録刊行物

• 繊維機械学会誌

  繊維機械学会誌 58 (6), T68-T75, 2005

  一般社団法人 日本繊維機械学会

キーワード

• 3D body shape model
• simulation of wearing a brassiere
• multiple regression model

詳細情報

• CRID: 1390001204265940352
• NII論文ID: 10014348176
• NII論文ID: 130004470562
• NII書誌ID: AN00285892
• DOI: 10.4188/transjtmsj.58.t68
• ISSN: 18801994
• ISSN: 03710580
• NDL書誌ID: 7408269
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I7408269
• 本文言語コード: ja
• データソース種別:
  • JaLC
  • NDL
  • Crossref
  • CiNii Articles
• 抄録ライセンスフラグ: 使用不可