基于三维高斯泼溅的三维人体重建方法及系统
By using a hierarchical hashing encoding method based on 3D Gaussian splashing and a joint loss function optimization method, the problems of computational efficiency and model accuracy in 3D human body reconstruction in monocular video are solved, and fast, high-fidelity 3D human body reconstruction is achieved.
CN122156415BActive Publication Date: 2026-07-17CHINA UNIV OF PETROLEUM (EAST CHINA)
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA UNIV OF PETROLEUM (EAST CHINA)
- Filing Date
- 2026-05-07
- Publication Date
- 2026-07-17
Smart Images
Figure CN122156415B_ABST
Abstract
本发明属于三维视觉与数字化技术领域，涉及一种基于三维高斯泼溅的三维人体重建方法及系统，方法步骤为：基于分层哈希编码参数场，校正构建的三维高斯基元集合中每个高斯基元的中心位置，并预测高斯基元在当前观察视角下的颜色；基于单目视频序列对应的人体姿态参数和形状参数，对校正后的高斯基元执行线性混合蒙皮变换，映射至姿态空间；基于映射至姿态空间的高斯基元参数，对映射至姿态空间的高斯基元进行三维高斯可微渲染，得到与单目视频对应视角一致的渲染图像；基于构建的联合损失函数，优化高斯基元参数与分层哈希编码参数场，得到三维人体模型。本发明能够实现从单目视频中快速重建轮廓稳定、纹理清晰的可动画化三维人体模型。
Need to check novelty before this filing date? Find Prior Art