位姿确定方法、装置、电子设备和可读存储介质
By using image processing and distance conversion technology from camera equipment, the problems of high cost and low accuracy of inertial navigation systems have been solved, achieving low-cost and high-precision pose estimation, which is suitable for autonomous driving.
CN116934849BActive Publication Date: 2026-07-17丰图科技(深圳)有限公司
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 丰图科技(深圳)有限公司
- Filing Date
- 2022-03-31
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Existing inertial navigation systems are expensive and have low accuracy in environments with poor GNSS signals, making them difficult to widely apply to vehicle pose estimation in autonomous driving.
Method used
By acquiring scene images and estimated poses captured by camera equipment, generating simulated images using preset maps and camera equipment parameters, and performing distance conversion, the pose error is determined to adjust the pose, thus achieving low-cost and high-precision pose estimation.
Benefits of technology
It improves pose accuracy at low cost, is suitable for environments with poor GNSS signals, and meets the needs of autonomous driving.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN116934849B_ABST
Abstract
本发明实施例提供一种位姿确定方法、装置、电子设备和可读存储介质，涉及图像处理技术领域。本发明实施例通过获取摄像设备拍摄的场景图像以及摄像设备对应的估计位姿，根据摄像设备的参数，确定得到预设地图中估计位姿对应的模拟图像，然后通过对场景图像进行距离转换，得到场景图像的距离转换图，然后根据模拟图像以及距离转换图，确定得到估计位姿下的误差，根据误差，确定摄像设备的目标位姿。如此，根据模拟图像以及距离转换图，确定得到估计位姿下的误差，根据误差，确定摄像设备的目标位姿，可以在保证低成本的情况下，有效提高位姿的精确度。
Need to check novelty before this filing date? Find Prior Art