Camera-Based Sensor Fusion for Low-Speed Vehicle Heading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic steering systems for vehicles face challenges in accurately determining vehicle heading, especially when the vehicle is stationary or operating at low speeds, due to gyro drift and poor GNSS signal quality.
Innovation Solution
The system integrates a camera system that uses visual odometry and SLAM (Simultaneous Localization and Mapping) to maintain accurate vehicle heading without drifting, combining this data with GNSS and IMU information using fusion algorithms such as complementary filters and Kalman filters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If wheel odometry, inertial navigation systems (INS) and degraded GNSS are used to solve heading issues, then vehicle control precision is improved, but system complexity and cost increase
Solution Approach 1:
The patent combines multiple sensor types (GNSS receiver, IMU, wheel speed sensors) into an integrated navigation system that fuses data from all sources through algorithms like EKF to achieve precise heading measurement without requiring complex dual-antenna systems or expensive specialized equipment
Solution Approach 2:
The system uses a single GNSS antenna that serves multiple functions: providing position data, velocity data through Doppler effect, and aiding in heading determination when combined with IMU data, eliminating the need for separate dedicated sensors for each measurement type
2Measurement precision
If dual antenna systems are used to measure heading and roll, then measurement precision is improved, but hardware cost and system complexity increase
Solution Approach 1:
The patent introduces an IMU as an intermediary device that measures vehicle orientation (roll, pitch, yaw) directly, providing heading information without requiring complex dual-antenna geometric calculations, thereby simplifying the antenna system to a single unit
Solution Approach 2:
The system replaces the mechanical/geometric dual-antenna heading measurement system with an inertial measurement system using gyroscopes and accelerometers in the IMU, which directly sense vehicle orientation through physical principles rather than geometric positioning
3Device complexity
If single antenna systems with gyros and accelerometers are used, then device complexity is reduced, but heading measurement precision deteriorates due to gyro drift
Solution Approach 1:
The system uses GNSS velocity data (derived from Doppler effect and position changes) as feedback to continuously correct and reset the drift accumulation in IMU integrated heading measurements, maintaining long-term accuracy without requiring complex dual-antenna systems
Solution Approach 2:
The system performs preliminary correction of IMU drift by periodically updating heading estimates using GNSS-derived course-over-ground information before significant drift accumulation occurs, preventing rather than correcting large errors
4Measurement precision
If GNSS course over ground information is used to compensate gyro bias, then measurement precision is improved, but reliability deteriorates at low speed and standstill due to noisy and delayed GNSS signals
Solution Approach 1:
The system dynamically adjusts the weighting and fusion strategy between IMU and GNSS data based on vehicle speed conditions, relying more on IMU predictions at low speeds where GNSS is noisy, and transitioning to GNSS-corrected fusion at higher speeds where GNSS data is reliable
Solution Approach 2:
The system changes the operational parameters of the sensor fusion algorithm based on vehicle speed, adjusting filter gains and data weighting to optimize performance across different speed regimes, ensuring reliable heading estimation whether the vehicle is stationary or moving at high speed
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A control system (100) fuses different sensor data together to determine an orientation of a vehicle (50). The control system (100) receives visual heading data for the vehicle (50) from a camera system (102), global navigation satellite system (GNSS) heading data from a GNSS system (108), and inertial measurement unit (IMU) heading data from an IMU (110). The control system (100) may assign weights to the visual, GNSS, and IMU heading data based on operating conditions of the vehicle (50) that can vary accuracy associated with the different visual, GNSS, and IMU data. The control system (100) then uses the weighted visual, GNSS, and IMU data to determine a more accurate vehicle heading.