Back to results list
Please use this identifier to cite or link to this item:
|Title:||Development of 2D block patterns from fit feature-aligned flattenable 3D garments||Authors:||Huang, Haiqiao||Keywords:||Dressmaking -- Pattern design.
Clothing and dress measurements -- Data processing.
Hong Kong Polytechnic University -- Dissertations
|Issue Date:||2011||Publisher:||The Hong Kong Polytechnic University||Abstract:||Patternmaking is a crucial step in apparel manufacturing. The dimensional fit of a garment largely depends on the quality of its clothing patterns. Block patterns, a special type of pattern, are widely used in the fashion industry as templates for pattern design. Diverse styles can be created from the blocks. The blocks thus determine the final garment fit. However, traditional block construction methods have some intrinsic drawbacks, and iterative trial fitting is necessary to improve the fit of the block patterns. Recently, 3D-to-2D computer modelling techniques have been applied to the garment industry. The technique can generate 2D planar shapes from 3D objects. A direct application of these techniques to pattern generation can be problematic for several reasons. Firstly, the shapes generated by these methods are either convex or free boundary that cannot be used as garment patterns. Secondly, the shape of the generated pattern is not fixed but varies due to different initial settings. More importantly, the 3D garments and 2D patterns created by the existing 3D-to-2D methods cannot fulfil the garment fit requirements, because neither the garment structural lines nor the necessary clothing ease has been considered in the pattern generation process. The aim of this study is to establish a systematic approach for 3D garments and 2D block patterns development. The structural lines of the 3D block garments are properly aligned with the features of the body. Garment fit is guaranteed by controlled ease distribution to the blocks in three-dimensional space. The generated block patterns should conform to the apparel manufacturing requirements. To achieve the stated objectives, a three-phase methodology is proposed and implemented in this thesis, which includes (1) human body data acquisition and parameterization, (2) feature alignment and wireframe establishment, and (3) flattenable 3D garments and 2D patterns development. In the first phase of the method, a scanned human body is processed to a recognisable model through noise filtering, basic body segmentation, Delaunay triangulation and model parameterization. In the second phase, twenty five body features are identified and extracted. A body feature wireframe is hence established, which is then purposely deformed to incorporate the desired garment eases to the wireframe. In the final phase, a developable surface approximation algorithm is used to construct garment surface patches based on the ease incorporated wireframe. 2D block patterns are then generated by flattening and rearranging 3D garment patches based on a graph-based algorithm. A wearing trial experiment has been conducted to verify the effectiveness of the proposed method in dealing with real world non-ideal body data. Eighteen subjects with different body shapes have participated in the wearing trial. The experimental results show that the proposed method out performs two conventional methods in development of customised block patterns with better fit. This research enhances the understanding of 3D garments and 2D patterns development. It realises efficient development of 2D block patterns from fit feature aligned and flattenable 3D garments.||Description:||xxviii, 297 p. : ill. (some col.) ; 30 cm.
PolyU Library Call No.: [THS] LG51 .H577P ITC 2011 Huang
|URI:||http://hdl.handle.net/10397/4951||Rights:||All rights reserved.|
|Appears in Collections:||Thesis|
Show full item record
Files in This Item:
|b24624949_link.htm||For PolyU Users||162 B||HTML||View/Open|
|b24624949_ir.pdf||For All Users (Non-printable)||5.45 MB||Adobe PDF||View/Open|
Citations as of Dec 10, 2018
Citations as of Dec 10, 2018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.