车端视角图像的生成方法、装置、电子设备及存储介质
By acquiring data from vehicle-mounted LiDAR and roadside sensors to generate static environment maps and fusing them with dynamic target data, and combining the motion trajectories of electronic devices and camera parameters for rendering, the problem of high cost of vehicle-mounted data acquisition and scarcity of complex scenes is solved, thereby improving the quality and realism of the generated images from the vehicle's perspective.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- TSINGHUA UNIVERSITY
- Filing Date
- 2026-03-23
- Publication Date
- 2026-07-17
AI Technical Summary
The high cost of vehicle-side data collection and annotation, coupled with the scarcity and difficulty in reproducing real-world, complex scenarios, presents significant challenges.
By acquiring vehicle-mounted LiDAR data and roadside sensor data of the target road, a static environment map is generated and fused with dynamic target data. Parametric projection rendering is then performed by combining the motion trajectory of electronic devices and camera installation parameters to generate a vehicle-view image.
It effectively reduced the cost of real vehicle data collection, improved the generation quality and realism of vehicle-view images, enhanced the fit between images and real vehicle imaging, and achieved the adaptation of images for various vehicle motion conditions.
Smart Images
Figure CN122415344A_ABST