Stereo Imagery Calibration Using Vehicle Motion and Static Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing stereo imagery calibration methods in vehicles, such as autonomous vehicles, face challenges in accurately correcting for dynamic changes due to factors like temperature, road conditions, and vibrations, leading to errors in depth estimation and yaw estimation, which are difficult to recover and can result in inaccurate navigation.
Innovation Solution
Utilizing motion-based disparity correction methods that leverage GPS and other satellite systems to determine the distance traveled and static feature disparities, applying corrections to disparity maps in real-time or substantially real-time, without significant computational overhead, by comparing the apparent motion of static features with known camera motion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional stereo imagery calibration methods are used, then the system structure remains simple, but measurement precision deteriorates due to dynamic changes from temperature, road conditions, and vibrations
Solution Approach 1:
The patent introduces GPS positioning data and inertial measurement unit (IMU) data as intermediary elements that mediate between the stereo camera system and the calibration process. These intermediaries provide independent motion reference information that compensates for dynamic changes in the stereo system caused by temperature, vibrations, and road conditions, thereby improving depth estimation accuracy without requiring complex hardware modifications to the stereo cameras themselves
Solution Approach 2:
The patent replaces traditional mechanical calibration methods (which rely on fixed physical calibration targets and stable mounting) with a computational approach using GPS and IMU data. Instead of mechanically ensuring stability through rigid mounting structures and fixed calibration targets, the system uses sensor fusion algorithms to computationally compensate for dynamic changes, substituting mechanical stability requirements with software-based correction
2Measurement precision
If motion-based disparity correction is applied in real-time, then navigation precision improves, but computational overhead increases
Solution Approach 1:
The patent performs preliminary action by pre-integrating GPS position data and IMU motion data to calculate expected camera motion before processing stereo imagery. The system pre-computes transformation matrices and motion compensation parameters based on recent sensor readings, so that when disparity correction is needed, the computationally intensive work has already been done, reducing real-time energy consumption while maintaining high yaw estimation accuracy
3Measurement precision
If GPS and satellite systems are integrated for motion tracking, then positioning accuracy improves, but device complexity increases
Solution Approach 1:
The patent applies universality by designing the sensor integration system to serve multiple functions simultaneously. The GPS and IMU sensors not only provide motion tracking data for disparity correction but also enable vehicle navigation, speed monitoring, and positioning functions. This multi-functionality justifies the increased device complexity by providing comprehensive vehicle monitoring capabilities beyond just stereo calibration
Solution Approach 2:
The patent merges GPS positioning data and IMU motion data into a unified motion reference framework that serves the stereo calibration system. By combining these independent sensor systems and fusing their data through sensor fusion algorithms, the patent creates a cohesive positioning and motion tracking solution that improves vehicle position accuracy while managing integration complexity through systematic data fusion
Data Source
AI summary
A system includes sensors onboard an autonomous vehicle, a processor, and a memory. The memory stores instructions for the processor to receive a first image pair from the sensors at a first time and a second image pair from the sensors at a second time. Each image pair includes at least one static feature in an environment of the autonomous vehicle. The memory also stores instructions to determine a distance travelled by the autonomous vehicle between the first and second times, and to determine a correction to a disparity map based on (1) a first disparity associated with the static feature(s) and the first image pair, (2) a second disparity associated with the static feature(s) and the second image pair, and (3) the distance travelled. The memory also stores instructions to cause the correction to be applied to the disparity map.