System development of a 3-dimensional fashion・factory・boutique(Part 1) Systematization and construction of the support informations of the 3D body shape, the feeling-of-wearing evaluation, and the design for women's dress selection support of the consumer

3维时装·工厂·精品店的系统开发(第1部分) 3D体型、穿着感评价、消费者选女装设计支持等支持信息的系统化和构建

• 批准号: 13558006
• 负责人: MASUDA Tomoe
• 金额: $ 8万
• 依托单位: Mie University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (B)
• 财政年份: 2001
• 资助国家: 日本
• 起止时间: 2001 至 2002
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-13558006/
• 关键词: Non-tactile 3D body measurement system; Woman; 3D body surface shape; Structured human body surface by triangularization; Concentrated Gaussian, Geodesic, and Mean curvatures; Vocabulary for design image of dress; Evaluation of image for dress desig

We try to obtain the information for selling & buying the dress of each individual body type and image for young women by means of using the 3D body measurment system and reserch of toe dress for sensuous image. The results were as follows.In cooperation with the manufacturer, it was inexpensive and the Non-tactile 3D body measurement system human body measuring instrument measurable in a short time was developed. For clothing design, the basic points (number of about 100) and lines (number of 6) and a marker's color, and the underwear for Non-tactile 3D body measurement were established. The 3D body shape of each model in 85 young women models is measurable in whole body about 10 seconds. A series of software to simulation of the 3D body shpe, the automatic formation by means of the structure of body surface using triangular mesh of a approximated torso body surface, and the calculation of the three curvatures (Gauss, geodesic, and means) was developed, and the systematization was att … More ained. However, some problems for the mesuring precision for the automatic extracting of a basic point etc. and the underwear for 3D mesurment remained. Therefore, only 3D data was usable for 58 models in 85 inside. The concentrated Gaussian, geodesic, and mean curvatures on the triangular vertices of 58 models' 3D torso surface (except for the side area) were computed automatically. The features of the surface shape were each area were investigated, and the body shape was also enabled visually confirm using the gradation display of the collar grade corresponding to a curvature value.For the young women, we researched the classification of the image on the vocabulary and the design of the dress for constructing the clothing select-support system. The image grouping (breakdown) of 243 words collected from the fashion magazine was done by 136 evaluator. As a result of cluster analysis, it had categorized into the typical image type (marine, ethnic, retro, casual, school, boys, hard, and garls, simple, and adult) of 10 words. On 60 design dress (photograph), the 50 evaluators judged for the image of the 10 words and the decisided design factors (a form, color, etc.) on the image of each dress using the metod of the 5-stage (step) evaluation. The 60 dress image were categorized into three groups using the cluster analysis. Discriminant analysis using the images and the design factors of each dress was performed. We deduced the formulas from the result, and the 3 images of the dress were able to distinct and estimate in 70%-90% of precision. The 3D fitting of the 20 dresses according as the high imaged design in previous evaluation were simulated on the supposition for the samples of the next dress image evaluation.From now on, we investigated the image evaluations and the classification using the many design dresses, and will evaluate of the image on the 3D fitting dress simulations and the 3D body shape. In this research, the fundamental products from many researches in different disciplines us were obtained for the structure of the 3D fashion design system. Less

利用三维人体测量系统，对鞋底服饰的感官形象进行研究，获取年轻女性每一种体型和形象的服装买卖信息。研究结果表明：与生产厂家合作，价格低廉，研制出了可短时间测量的非触觉三维人体测量仪。对于服装设计，建立了基本点(约100个)和线条(6个)和一个记号的颜色，并建立了非触觉3D人体测量的内衣。85名年轻女性模特的3D体型在10秒左右就可以在全身范围内测量到。开发了一套三维体轮廓模拟、近似躯干体面三角网格自动生成、三种曲率(高斯曲率、测地线曲率、均值曲率)的计算软件，并在…上实现了系统化更加无精打采。然而，在基点自动提取等测量精度和三维测量内衣等方面仍存在一些问题。因此，只有3D数据可用于85个内部的58个模型。自动计算了58个模型的三维躯干曲面(侧面除外)三角顶点上的集中高斯曲率、测地线曲率和平均曲率。针对年轻女性，研究了基于词汇的图像分类和服装设计，构建了服装选配支持系统。从时尚杂志收集的243个单词的图像分组(分解)是由136名评估者完成的。聚类分析的结果是，它被归类为10个单词的典型意象类型(海洋、种族、复古、休闲、学校、男孩、硬汉、花环、简单和成人)。在60件设计服装(照片)上，50名评价者根据10个单词的形象和决定性的设计因素(形式、颜色等)进行评判。对每件衣服的形象采用了5阶段(步骤)的评价方法。通过聚类分析，将60个服装图像分为三组。对每件衣服的图像和设计因素进行了判别分析。根据实验结果推导出计算公式，3幅服装图像的识别精度为70%-90%。根据以往评价中的高形象设计对20件礼服的3D试衣进行了模拟，为下一次的服装形象评价提供了样本。从现在开始，我们研究了利用大量设计的礼服进行形象评价和分类，并将对图像进行3D试衣模拟和3D体型评价。在本研究中，我们从不同学科的多项研究中获得了构建三维服装设计系统的基础产品。较少

项目成果

Tomoe Mauda, Haruki Imaoka: "Feature Examination of Curved Shapes for Tight-Fitting Patterns of Young Women by Using Concentrated Gaussian Curvature"J.Home Econ.Jpn.. 53, No.8. 795-804 (2002)

Tomoe Mauda、Haruki Imaoka：“使用集中高斯曲率对年轻女性紧身图案的曲线形状进行特征检查”J.Home Econ.Jpn.. 53，第 8 期。

T.Masuda, H.Imaoka: "Feature Examination of Curved Shapes for Tight-Fitting Patterns of Young Women by Using Concentrated Gaussian Curvature"J. Home Econ. Jpn.. 53・8. 795-804 (2002)

T.Masuda、H.Imaoka：“使用集中高斯曲率对年轻女性紧身图案的曲线形状进行特征检查”J. Home Econ. 53・8 (2002)。