VR blood vessel real-time display method and device based on low-delay virtual-real fusion
By combining lightweight neural networks and perspective multi-point localization algorithms with low-latency transmission and fusion rendering optimization, the pose alignment and surface fusion problems in VR blood vessel display methods are solved, achieving high-quality virtual-real fusion display.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PLASTIC SURGERY HOSPITAL CHINESE ACADEMY OF MEDICAL SCIENCES
- Filing Date
- 2026-03-16
- Publication Date
- 2026-06-26
AI Technical Summary
Existing VR blood vessel display methods have shortcomings in pose alignment, low-latency transmission, and curved surface blending rendering, which affect the spatial alignment accuracy and display stability of virtual-real fusion, making it difficult to achieve high-quality pixel-level virtual-real blending.
Accurate real-time solution of head pose transformation matrix is achieved by using lightweight neural network feature point extraction and perspective multi-point localization algorithm. A low-latency transmission mechanism is constructed, which combines row buffer triggered piecewise encoding and pose deviation affine distortion correction. The fusion rendering is performed by combining thin plate spline surface adaptation and stage adaptive transparency blending.
It effectively solves the shortcomings of traditional technologies in pose alignment, low-latency transmission, and curved surface fusion rendering, ensuring the technical guarantee for real-time visualization of VR blood vessels and realizing high-quality virtual-real fusion display.
Smart Images

Figure CN122289615A_ABST