Stereo Vision Yaw Error Calibration Using Non-Linear Solvers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vision systems for motor vehicles face challenges in accurately determining the yaw angle error between stereo cameras due to thermal changes over time, requiring online solutions that do not rely on external sensors like radar or lidar, and are prone to errors from odometric data uncertainties.
Innovation Solution
A non-linear equation solver method is used to estimate the intrinsic yaw error by relating time frames, calculated disparity values, and kinematic variables of the ego vehicle, with a set of equations that express the current distance to detected objects, incorporating the yaw error as a shift to the disparity value, allowing for accurate calibration during vehicle movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radar or lidar reference systems are used to estimate squint angle error, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The vision system uses its own stereo camera images and internal processing capabilities to determine squint angle error, without requiring external radar or lidar reference systems. The system processes image data from the stereo cameras themselves to calculate disparity values and derive the squint angle error, making the system self-sufficient and avoiding additional sensor hardware.
2Productivity
If external odometric data is used to calculate reference driven distance, then productivity is improved, but measurement precision deteriorates due to systematic uncertainty
Solution Approach 1:
The method extracts and eliminates the dependency on external odometric data by using only image data from the stereo cameras themselves. The squint angle error is determined directly from disparity calculations on stereo images without incorporating external reference data, thereby removing the source of systematic uncertainty while maintaining calibration efficiency through online processing.
3Measurement precision
If exact knowledge of ego vehicle movement from other sensors is required, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The vision system determines squint angle error using only the stereo camera images and internal processing, without requiring external sensors to provide ego vehicle movement data. The method calculates disparity values and derives the squint angle error from image data alone, making the system self-sufficient and avoiding additional sensor hardware requirements.
4Adaptability or versatility
If online estimation of squint angle error is implemented, then adaptability is improved, but computational effort increases
Solution Approach 1:
The system continuously processes stereo images to maintain up-to-date squint angle error estimation during vehicle operation. By processing image data in real-time as it is captured, the system maintains adaptability to thermal changes and operational conditions without requiring computationally intensive batch processing or offline calibration procedures.
Data Source
AI summary
A vision system (10) for a motor vehicle with a stereo imaging apparatus (11) with imaging devices (12) adapted to capture images from a surrounding of the motor vehicle, and a processing device (14) adapted to process images captured by the imaging devices (12) and to detect objects, and track detected objects over several time frames, in the captured images. The processing device (14) is adapted to obtain an estimated value for the intrinsic yaw error of the imaging devices (12) by solving a set of equations, belonging to one particular detected object (30), using a non-linear equation solver method, where each equation corresponds to one time frame and relates a frame time, a disparity value of the particular detected object, an intrinsic yaw error and a kinematic variable of the vehicle.