Sensor Fusion Distance Normalization for Object Association
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensor fusion systems face challenges in achieving highly reliable object association in a short period due to issues with Euclidean distance calculations and fixed thresholds, leading to erroneous associations or prolonged processing times.
Innovation Solution
A sensor information integration system that includes a distance calculation part normalizing distances based on directional axis preciseness information and an association determination part using post-normalization distances to accurately match objects between sensors, dynamically adjusting thresholds for improved reliability and speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If Euclidean distance calculation with fixed threshold is used for object association, then the association processing is simple, but the reliability of association deteriorates due to erroneous associations or missed associations
Solution Approach 1:
The patent applies local quality by evaluating distance along each directional axis separately with its own precision-based threshold, rather than using a single uniform Euclidean distance threshold. Each axis (e.g., longitudinal and lateral directions) has its threshold adjusted according to the specific precision characteristics of sensors in that direction, allowing the association process to account for directional variations in measurement quality.
Solution Approach 2:
The patent implements dynamics by making the association threshold dynamic rather than fixed. The threshold for each directional axis is determined based on the precision information of the sensors, which can vary depending on sensor type, distance, and environmental conditions. This dynamic threshold adaptation allows the system to maintain high association reliability across different operating conditions.
2Productivity
If fixed threshold is used for object association, then the processing speed is fast, but the precision of association deteriorates leading to erroneous detections
Solution Approach 1:
The patent applies parameter changes by modifying the association threshold parameter based on precision information. Instead of using a constant threshold, the system calculates dynamic thresholds for each directional axis by incorporating precision values from sensor data. This allows the threshold parameter to adapt to the actual measurement quality, improving association precision while maintaining processing efficiency through automated calculation.
3Device complexity
If Euclidean distance is used without considering directional precision, then the calculation is simple, but the reliability of distance-based association deteriorates
Solution Approach 1:
The patent applies segmentation by dividing the distance calculation into separate directional components (e.g., longitudinal and lateral axes). Instead of calculating a single Euclidean distance, the system computes distances along each axis independently and evaluates them separately against direction-specific thresholds. This segmentation allows the system to account for different precision characteristics in different directions while keeping the overall process manageable.
Solution Approach 2:
The patent applies local quality by assigning different quality weights to different directional axes based on sensor precision. Each directional axis is evaluated with its own precision-adjusted threshold, reflecting the local measurement quality in that direction. This approach ensures that distance-based association relies more heavily on directions with higher precision and compensates for directions with lower precision.
Data Source
AI summary
A system includes a first analysis apparatus, a second analysis apparatus, and a sensor fusion apparatus. The individual analysis apparatuses analyze outputs of their respective sensors and generate preciseness information about preciseness of the positions of their respective objects. The sensor fusion apparatus calculates the distance between a determination target object and an association candidate object per directional axis that is determined from a position of a sensor corresponding to the determination target object. The sensor fusion apparatus normalizes the distance calculated per directional axis by using the preciseness information corresponding to the determination target object. The sensor fusion apparatus calculates the distance between the determination target object and the association candidate object by using the distance normalized per directional axis as a post-normalization distance. The sensor fusion apparatus determines whether the association candidate object matches the determination target object by using the post-normalization distance.