Visual Odometry for Vehicle Position Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for generating a bird's-eye view using three-dimensional spatial information from two-dimensional images captured by vehicle-mounted cameras are prone to synchronization errors due to communication delays on the CAN bus, which affect the accuracy of vehicle position and orientation estimation.
Innovation Solution
The method employs visual odometry based on sequential two-dimensional images captured by vehicle-mounted cameras to estimate vehicle position and orientation, and corrects bird's-eye view features using this information, reducing the influence of synchronization errors by avoiding communication delays typically associated with sensors connected via the CAN bus.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensors connected via CAN bus are used to obtain vehicle motion information, then the system can access acceleration, speed, and yaw rate data, but communication delays cause synchronization errors that reduce measurement precision
Solution Approach 1:
The patent introduces an image processing-based visual odometry system as an intermediary to obtain vehicle motion information. Instead of directly using sensor data from the CAN bus, the system processes sequential images to calculate vehicle position and orientation, thereby avoiding the synchronization errors inherent in CAN bus communication while still obtaining the necessary motion parameters for BEV feature correction.
2Measurement precision
If visual odometry based on image processing is used to estimate vehicle position and orientation, then synchronization errors are reduced, but the computational complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-calculating and storing the extrinsic parameters (rotation and translation matrices) between the camera coordinate system and vehicle coordinate system. This pre-processing step simplifies the main computation by allowing direct application of these predetermined transformation matrices to correct BEV features, rather than performing complex coordinate transformations in real-time.
3Measurement precision
If BEV features are corrected using visual odometry information, then the accuracy of bird's-eye view generation is improved, but the processing time increases
Solution Approach 1:
The patent replaces the traditional sensor-based mechanical information acquisition system with an optical-based visual odometry system. By substituting physical sensor measurements with image processing-derived motion estimation, the system achieves better synchronization accuracy while maintaining real-time performance through efficient computational methods.
Data Source
AI summary
An estimation device is configured to: calculate a position and an orientation of a vehicle, using a self-position estimation method including Visual Odometry, based on sequential two-dimensional images of outside of the vehicle captured by a same camera, which is at least one of a plurality of cameras provided on the vehicle; obtain a bird's-eye view (BEV) feature, which is a feature in a BEV space, based on the two-dimensional images of the outside of the vehicle using a BEV estimation algorithm; and correct the obtained BEV feature using information representing the position and the orientation.