Multi-Sensor Axis Correction Using Road Feature Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driving assistance and automatic driving systems face challenges in determining sensor axis deviation when the detection areas of multiple sensors do not overlap, as they rely on overlapping areas for accurate axis correction.
Innovation Solution
A sensor fusion device that includes a sensor coordinate conversion unit, a target selection unit, a function fitting unit, and a correction value calculation unit to convert and correct sensor data from multiple sensors into a unified coordinate system, allowing for axis deviation correction even when detection areas do not overlap.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor axis deviation correction is performed using overlapping detection areas, then measurement precision is improved, but the system cannot determine axis deviation when detection areas do not overlap
Solution Approach 1:
The patent introduces road surface features (white lines, lane markings, curbs) as intermediary objects that serve as common reference points between sensors. These features act as mediators that allow the system to calculate relative positions and determine axis deviation even when sensor detection areas do not directly overlap, by providing a shared coordinate framework based on the road surface.
Solution Approach 2:
The patent creates a virtual model (copy) of the road surface features detected by one sensor and compares it with the actual detection results of another sensor. By generating expected detection positions based on the known geometry of road markings and comparing these with actual sensor readings, the system can determine axis deviation without requiring direct detection area overlap.
2Area of stationary object
If multiple sensors are used for comprehensive monitoring, then coverage area is improved, but determining axis deviation becomes more difficult when detection areas do not overlap
Solution Approach 1:
Road surface features serve as intermediary reference objects that connect multiple sensors' detection spaces. By using these common road features as mediators, the system can establish coordinate relationships between sensors with non-overlapping detection areas, making axis deviation determination feasible across the entire detection coverage area.
Solution Approach 2:
The patent changes the approach from direct sensor-to-sensor comparison to sensor-to-road-feature comparison. By introducing road surface feature parameters (position, orientation, geometry) as intermediate variables, the system transforms the complex multi-sensor coordination problem into simpler individual sensor calibration problems relative to the road coordinate system.
Data Source
AI summary
To correct an axis deviation of a sensor. An aiming device, which calculates correction amounts of detection results of two or more sensors using the detection results of the sensors, includes: a sensor coordinate conversion unit that converts sensor data detected by the sensor from a coordinate system unique to the sensor into a predetermined unified coordinate system; a target selection unit that selects predetermined features from the sensor data detected by each of the sensors; a function fitting unit that defines functions each approximating an array state of the selected features for the respective sensors; a fitting result comparison unit that compares the functions each approximating the array state of the features detected by each of the sensors; and a correction value calculation unit that calculates a correction amount for converting coordinates of the features detected by the sensors from a result of the comparison of the functions.