计算机与现代化 ›› 2025, Vol. 0 ›› Issue (09): 90-96.doi: 10.3969/j.issn.1006-2475.2025.09.013

• 图像处理 • 上一篇    下一篇

基于Mask约束的单视图3D头发建模

  


  1. (安徽理工大学人工智能学院,安徽 淮南 232000)

  • 出版日期:2025-09-24 发布日期:2025-09-24
  • 作者简介: 作者简介:李谡(2000 —),男,安徽淮北人,硕士研究生,研究方向:计算机视觉,图像处理,E-mail: leesucs@aust.edu.cn; 通信作者:李德权(1973—),男,安徽合肥人,教授,博士,研究方向:人工智能算法,计算机视觉,机器人智能控制,E-mail:dqli@aust.edu.cn; 庆宇(1999—),男,安徽马鞍山人,硕士研究生,研究方向:信号处理,图像处理,E-mail: 2022201831@aust.edu.cn。
  • 基金资助:
     基金项目:安徽省自然科学基金资助项目(2208085ME128); 安徽省学术和技术带头人及后备人选项目(2019h211)
       

Single View 3D Hair Modeling Based on Mask Constraints


  1. (School of Artificial Intelligence, Anhui University of Science & Technology, Huainan 232000, China)
  • Online:2025-09-24 Published:2025-09-24

摘要: 摘要:提出一种基于Mask约束的3D头发重建方法。首先,利用SAM推理出头发正视Mask,并通过GAN网络预测合理的后视Mask,引入深度伪标签微调Hourglass网络预测头发深度图,增强头发的相对位置关系,另外使用U-Net推理头发的方向图,将这些信息输入Stacked Hourglass网络,并且分别利用正视Mask和后视Mask约束生成三维头发空间点,确保生成的头发三维点云在Mask范围内投影合理。同时,通过控制负样本采样率,增强模型在头发边缘的鲁棒性。最后,采用并行算法加速头发合成,显著提高重建效率。实验结果表明,该方法在处理复杂发型和提高头发真实感方面效果显著。


关键词: 关键词:头发建模, 基于图像的头发建模, 占用场, 三维重建, 方向场

Abstract:
Abstract: A 3D hair reconstruction method based on Mask constraints is proposed. Firstly, the hair front view Mask is inferred using SAM, and a reasonable back view Mask is predicted by GAN network, the depth pseudo-labeling is introduced to fine-tune the Hourglass network to predict the hair depth map and enhance the relative positional relationship of the hair, and in addition, the U-Net is used to infer the hair orientation map, which is inputted into the Stacked Hourglass network, and respectively the front-view Mask and back-view Mask constraints to generate 3D hair spatial points to ensure that the generated 3D point cloud of hair is projected reasonably within the Mask range. Meanwhile, the robustness of the model at the hair edge is enhanced by controlling the negative sample sampling rate. Finally, a parallel algorithm is used to accelerate the hair synthesis and significantly improve the reconstruction efficiency. The experimental results show that the method is effective in dealing with complex hairstyles and improving the realism of hair.

Key words: Key words: hair modeling, image-based hair modeling, occupation field, 3D reconstruction, direction field

中图分类号: