Vehicle Sensor Fusion with Uncertainty-Based Reference Point Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensor data fusion methods for vehicles face challenges in reliably assigning and fusing sensor and fusion objects due to differences in measuring principles and limited computing power, leading to incorrect reference point information and compromised fusion results.
Innovation Solution
A method that determines reference point transformation candidates based on uncertainty indicators, such as variances and sensor capabilities, to transform sensor object data into fusion object data, ensuring accurate association and fusion by restricting transformations according to known quantities and sensor capabilities, and evaluating objects as point targets when uncertainties exceed thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If high-level sensor data fusion is used to reduce computational load, then computing power requirements are reduced, but measurement precision and reliability of object representation deteriorate due to loss of detailed reference point information
Solution Approach 1:
The patent segments the fusion process into two stages: first fusing sensor data to create initial fusion objects with abstract reference points, then performing a second fusion stage that specifically processes reference point transformations. This segmentation allows the system to maintain detailed reference point information when needed while keeping overall computational load manageable through the abstract initial fusion.
Solution Approach 2:
The patent performs preliminary determination of reference point transformation candidates before the actual fusion operation. By pre-identifying which reference points can be transformed and which cannot (based on indicator data about determination uncertainty), the system prepares the data structure in advance, reducing the computational complexity of the subsequent fusion operation while preserving measurement precision.
2Measurement precision
If reference point transformation is performed to align sensor objects with fusion objects, then object association accuracy is improved, but device complexity increases due to additional transformation calculations
Solution Approach 1:
The patent applies local quality by treating different reference points differently based on their transformation capabilities. Reference points are categorized into transformation candidates (where transformation is applicable) and non-candidates (where it is not). This local differentiation allows the system to apply transformation calculations only where beneficial, improving object association accuracy without unnecessarily increasing device complexity across all reference points.
Solution Approach 2:
The patent performs partial transformation actions by selectively applying reference point transformations only to transformation candidates rather than all reference points. The indicator data determines which transformations are performed, allowing the system to achieve sufficient object association accuracy without the excessive computational complexity of universal transformation.
3Reliability
If all reference points are transformed to achieve accurate fusion, then fusion reliability is improved, but loss of time increases due to extensive transformation calculations
Solution Approach 1:
The patent performs only the necessary partial transformations by using indicator data to identify transformation candidates. Instead of transforming all reference points to ensure fusion reliability, the system transforms only those reference points where transformation is determined to be beneficial and feasible, thereby maintaining fusion reliability while significantly reducing the time loss associated with unnecessary transformation calculations.
4Manufacturing precision
If detailed reference point treatment is applied to all sensor objects, then manufacturing precision of object representation is improved, but productivity of sensor data fusion decreases due to increased processing requirements
Solution Approach 1:
The patent applies local quality by providing detailed reference point treatment only for transformation candidates rather than all sensor objects. The indicator data enables the system to identify which objects require detailed reference point processing, thereby maintaining high object representation precision for critical cases while improving overall sensor data fusion productivity by avoiding unnecessary detailed processing of other objects.
Data Source
AI summary
A method and device for sensor data fusion for a vehicle as well as a computer program and a computer-readable storage medium are disclosed. At least one sensor device (S1) is associated with the vehicle (F), and in the method, fusion object data is provided representative of a fusion object (OF) detected in an environment of the vehicle (F); sensor object data is provided representative of a sensor object (OS) detected by the sensor device (S1) in the environment of the vehicle (F); indicator data is provided representative of an uncertainty in the determination of the sensor object data; reference point transformation candidates of the sensor object (OS) are determined depending on the indicator data; and an innovated fusion object is determined depending on the reference point transformation candidates.