纺织学报 ›› 2024, Vol. 45 ›› Issue (11): 80-87.doi: 10.13475/j.fzxb.20230806801

• 纺织工程 • 上一篇    下一篇

基于点云的三维纱线参数化建模

李文雅(), 梁建行, 薛涛, 董真真   

  1. 西安工程大学 纺织科学与工程学院, 陕西 西安 710048
  • 收稿日期:2023-08-18 修回日期:2024-04-20 出版日期:2024-11-15 发布日期:2024-12-30
  • 作者简介:李文雅(1988—),女,副教授,博士。研究方向为纺织计算机软件的开发及产品仿真。E-mail:leewya@126.com
  • 基金资助:
    陕西省教育厅重点科学研究计划项目(20JY026);中国纺织工业联合会科技指导性项目(2022020);博士科研基金项目(BS202105);陕西省科技计划陕西省重点研发项目一般项目(2024GX-YBXM-569)

Three-dimensional parametric modeling of yarn based on point clouds

LI Wenya(), LIANG Jianhang, XUE Tao, DONG Zhenzhen   

  1. School of Textile Science and Engineering, Xi'an Polytechnic University, Xi'an, Shaanxi 710048, China
  • Received:2023-08-18 Revised:2024-04-20 Published:2024-11-15 Online:2024-12-30

摘要:

针对纱线生产和用于织物开发中的设计周期长、打样成本高等问题,提出了基于点云的方法来进行三维纱线参数化建模,构建了短纤维纱线(如纯棉纱)的模型。首先以空间圆参数方程和Python的图形库Circle为基点建立圆堆图,然后结合不同象限的螺旋线捻向公式,以圆堆图为起点建立空间样条曲线。根据纱线中纤维的扭曲特性,将空间曲线划分为多个等距线段,并以这些线段的切线方向作为近似法向量,进而构建出由三维点云表示的多个空间圆,然后利用点云曲面重建算法将点云构建为初始三角网格模型,最后对初始三角网格模型进行曲面细分,从而得到光滑且精度高的三维纱线模型。实验结果表明,该方法能快速生成真实的三维纱线模型,且模型中包含纤维根数、纤维直径、纱线捻向、毛羽等多个特征点,可提高生产效率。

关键词: 圆堆图, 空间曲线, 点云, 三角网格, 纱线模型

Abstract:

Objective The three-dimensional yarn simulation is an important part for yarn design and fabric development. The use of computer simulation technology to simulate and visualize the yarn appearance is the key part of the current research. At present, most yarn simulations are carried out in two-dimensional or three-dimensional space, so that the appearance and shape of the yarn are limited. Therefore, it is necessary to design a fast and real three -dimensional yarn modeling method.

Method Circle stack diagram is first established based on the space circle and the Circle library, and then the spatial spline curve is established by combining the spiral twist equations of the circle stack diagram in different quadrants. According to the twisting characteristics of the fibers in the yarn, the curve is divided into multiple equidistant line segments, the spatial circle represented by the 3-D point cloud is established with the line segments as the normal vector, and the point cloud meshing algorithm is used to construct the point cloud as the initial triangular mesh model. The Loop subdivision algorithm is used to subdivide the mesh model.

Results The effectiveness of the 3-D yarn model based on point cloud method is verified and analyzed, and the 3-D yarn model established by different methods was compared and demonstrated. According to the effect of the 3-D yarn model established in literature, the geometric tube is modeled according to the input yarn parameters to match these fibers, and several single yarns with different structures are simulated, which shows that this is an effective yarn modeling and rendering method. In the three-dimensional yarn model based on the point cloud method, the distribution of fibers in the yarn cross-section is uneven, therefore it is necessary to deal with the regular circle pile diagram established by the regular yarn cross-section. Through the randomness of the starting point of fiber generation and the twisting state of the fiber, combined with the spiral twisting formula of different quadrants, the three-dimensional spatial spline without splicing state can be continuously generated. According to the twisting characteristics of the fibers in the yarn, the curve is divided into multiple equidistant segments, and the spatial circle represented by the 3-D point cloud is established with the line segments as the normal vector. Then, the point cloud grid algorithm is used to construct the point cloud as an initial triangular mesh model. Finally, the Loop subdivision algorithm is used to subdivide the surface of the mesh model, and the number of subdivisions is set according to the required model accuracy to improve the overall smoothness of the 3-D yarn model. In addition, the hairiness on the 3-D yarn model is formed by one end of the fiber protruding from the surface of the yarn body, which is consistent with the generation state of the actual yarn hairiness. Experiments show that the method can quickly generate a real 3-D yarn model, and the model contains multiple feature points such as number of fibers, fiber diameter, yarn twist direction, hairiness, and so on, which further illustrates the effectiveness of the method.

Conclusion Based on the importance of 3-D yarn simulation in yarn design and fabric development, a method for parametric modelling of 3-D yarn based on point cloud is proposed. The starting point of 3-D yarn model establishment is the circle stack diagram, the spatial spline curve composed of multiple equidistant line segments is created according to the custom conditions, and the spatial scattering circle expressed by 3-D point cloud is generated with the line segments as the center of the circle, and then the 3-D point cloud reconstruction algorithm is used to construct the point cloud between multiple spatial scattering circles into a triangular mesh model, and finally the surface subdivision of the initial 3-D yarn model is carried out to obtain the 3-D yarn model with high accuracy. The experimental results show that the 3-D visual effect of the yarn model is more obvious based on the number of constructed point clouds and the optimization of subdivision. At the same time, the algorithm realistically expresses the geometric characteristics of the simulated object, such as yarn thickness, hairiness, twist direction and fiber diameter, and so on, and can be adjusted according to the actual situation. The simulation is easy to operate and runs fast, which can better assist yarn design and development for practice.

Key words: circle packing, spatial curve, point cloud, triangular grid, yarn model

中图分类号: 

  • TS101.8

图1

圆形嵌套布局算法实现圆堆图"

图2

空间样条曲线"

图3

不同扭曲与不同周期状态的样条线"

图4

点云模型"

图5

三维点云重建网格过程"

图6

面列表示例"

图7

基于点云曲线的三角网格图"

图8

网格定义颜色"

图9

纱线模型构建顺序"

图10

模型曲面细分流程"

图11

三维纱线模型"

表1

纺纱工艺参数表"

图片
编号
纤维颜色
(RGB)
混合比 纱线
类别
捻度/
(捻·m-1)
捻向
图12(a) (107,190,140)
(107,182,151)
(168,207,190)
(235,235,235)
0.7
0.1
0.1
0.1
混色纱 850 Z
图12(b) (107,28,90)
(137,58,140)
(149,65,128)
(235,235,235)
0.85
0.05
0.05
0.05
混色纱 850 Z
图12(c) (113,45,111)
(115,176,142)
0.5
0.5
赛络纺纱 850 Z

图12

实际纺制的短纤维纱线与三维短纤维纱线模型对比图"

[1] 韩要宾, 张杰, 潘如如, 等. 纤维尺度下数码纱仿真方法[J]. 纺织学报, 2017, 38(11): 150-155.
doi: 10.13475/j.fzxb.20160708606
HAN Yaobin, ZHANG Jie, PAN Ruru, et al. Simulation method of digital yarn at fiber scale[J]. Journal of Textile Research, 2017, 38(11): 150-155.
[2] KARGAR M, PAYVANDY P, MOZAFARY V. A numerical structure-based method of simulating filament yarns[J]. The Journal of the Textile Institute, 2022, 113(12): 2681-2693.
[3] LIU H, KYOSEV Y, JIANG G, et al. Realistic fabric rendering with yarn models[J]. Textile Research Journal, 2023, 93(15/16): 3552-3563.
[4] ZHAN S, LUAN F, BALA K. Fitting procedural yarn models for realistic cloth rendering[J]. ACM Transactions on Graphics, 2016, 35(4) : 1-11.
[5] WANG J, AKANKWASA NT, ZHANG Y, et al. The three-dimensional model of staple yarn formation simulated from randomly assembled fibers[J]. Textile Research Journal, 2020, 90(7/8): 866-876.
[6] 郑成杰, 郑之. 基于最大分类器差异域适应方法的三维点云分类[J]. 信息与控制, 2023, 52(5): 588-597.
ZHENG Chengjie, ZHENG Zhi. 3 dimensions based on the maxi mum classifier difference domain adaptation method point cloud classification[J]. Information and Control, 2023, 52(5): 588-597.
[7] 王文曦, 李乐林. 深度学习在点云分类中的研究综述[J]. 计算机工程与应用, 2022, 58(1): 26-40.
doi: 10.3778/j.issn.1002-8331.2105-0200
WANG Wenxi, LI Lelin. Review of deep learning in point cloud classification[J]. Computer Engineering and Applications, 2022, 58(1): 26-40.
doi: 10.3778/j.issn.1002-8331.2105-0200
[8] CHEN Y, LIU J, NI B, et al. Shape self-correction for unsupervised point cloud understanding[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 8382-8391.
[9] CHERCHI G, LIVESU M, SCATENI R, et al. Fast and robust mesh arrangements using floating-point arithmetic[J]. ACM Transactions on Graphics (TOG), 2020, 39(6): 1-16.
[10] LOOP C. Smooth subdivision surfaces based on triangles[D]. Utah: Masters Thesis University, 1987: 2-31.
[1] 金守峰, 沈文军, 肖福礼, 李毅. 基于线结构光的钢领内表面圆度测量方法[J]. 纺织学报, 2023, 44(10): 164-171.
[2] 潘博, 钟跃崎. 基于二维图像的三维服装重建[J]. 纺织学报, 2020, 41(04): 123-128.
[3] 顾冰菲, 闫彦红, 苏军强, 刘国联. 女西装特征部位松量分布规则[J]. 纺织学报, 2019, 40(05): 107-112.
[4] 鲍陈, 缪永伟, 孙瑜亮, 张旭东. 基于散乱点云的三维人体自动测量[J]. 纺织学报, 2019, 40(01): 120-129.
[5] 高晓晓 江红霞. 应用3D打印技术的运动文胸模杯个性化定制[J]. 纺织学报, 2018, 39(11): 135-139.
[6] 汝欣 彭来湖 吕明来 史伟民 胡旭东. 纬编针织物几何建模及其算法[J]. 纺织学报, 2018, 39(09): 44-49.
[7] 孟想 辛斌杰 李佳平. 采用多角度成像技术的纺织品三维轮廓重建算法[J]. 纺织学报, 2018, 39(04): 144-150.
[8] 朱广舟 李晓久. 人体扫描线点云精简与孔洞修补[J]. 纺织学报, 2015, 36(03): 110-114.
[9] 李祖华. 服装设计用三维人体扫描系统[J]. 纺织学报, 2014, 35(1): 144-.
[10] 李晓志 李晓久. 基于部位划分的三维人体曲面模型的构建[J]. 纺织学报, 2013, 34(9): 145-0.
[11] 李晓久, 景晓宁. 基于非接触式测量的人体点云简化方法[J]. 纺织学报, 2012, 33(7): 144-148.
[12] 李晓志;李晓久;王玉秀. 基于人体扫描线点云的光顺去噪[J]. 纺织学报, 2010, 31(3): 133-138.
[13] 刘军;李重;胡觉亮. 基于特征匹配的三维试衣效果[J]. 纺织学报, 2009, 30(01): 122-126.
[14] 李燕;黄凯. 基于Geomagic的三维人体建模技术[J]. 纺织学报, 2008, 29(5): 130-134.
[15] 张瑞云;黄新林;李汝勤. 机织物的计算机三维模拟[J]. 纺织学报, 2005, 26(1): 62-63.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!