一种三维虚拟服饰的绘制方法及相关装置
By dynamically adjusting the rendering precision of virtual clothing by combining eye movement, head movement, gestures, and semantic data, the problem of high computational overhead in virtual reality technology is solved, achieving a balance between visual realism and system performance, and improving the rendering efficiency and smoothness of virtual clothing.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HUNAN HAPPLY SUNSHINE INTERACTIVE ENTERTAINMENT MEDIA CO LTD
- Filing Date
- 2026-03-27
- Publication Date
- 2026-07-17
AI Technical Summary
In existing virtual reality technologies, the high-precision physical simulation of 3D clothing leads to high computational overhead, affecting the real-time performance and smoothness of the system, making it difficult to achieve a balance between visual realism and high system performance.
By integrating eye-tracking, head-tracking, gesture, and semantic data, the rendering precision of virtual clothing is dynamically adjusted. Computational resources are concentrated on the user's attention area. Multi-dimensional data is used to determine the user's interest value and perform high-precision rendering. The gaze area is determined by combining a dual constraint method of 3D space and UV space, and physical simulation algorithms of different precision are used for rendering.
It achieves high-fidelity clothing physics simulation while maintaining visual realism and reducing computational overhead, improving system fluency and performance.
Smart Images
Figure CN121921475B_ABST