Vehicle Sensor Confidence Index Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining a confidence index for objects detected by sensors in motor vehicles do not accurately represent the variation in detection quality, leading to either missing real objects or incorrectly identifying non-existent ones due to a uniform threshold approach.
Innovation Solution
A method that approximates the detection zone as a polygon, determines the center of gravity, divides it into triangular sub-zones, identifies the sub-zone containing the object, and calculates a confidence index using a decreasing function based on the homothetic ratio between the object's line and the sub-zone's outer border, providing a more precise representation of detection quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a uniform threshold approach is used for confidence index, then the system is simple to operate, but detection precision deteriorates due to inability to represent variation in detection quality
Solution Approach 1:
The detection zone is segmented into multiple triangular sub-zones radiating from the sensor, allowing different confidence calculations for different spatial regions. This segmentation enables the system to account for varying detection qualities across the field of view without requiring a completely complex overall system.
Solution Approach 2:
Each triangular sub-zone is assigned its own confidence index calculation based on local geometric properties (distance from sensor, angular position). This local quality approach allows the system to adapt confidence thresholds to specific regions of the detection zone, improving overall detection precision while maintaining relatively simple local calculations.
2Reliability
If a low threshold percentage is used to avoid missing real objects, then detection sensitivity improves, but false positives increase due to inclusion of non-existent objects
Solution Approach 1:
The confidence index calculation uses multiple parameters including distance from sensor, angular position, and triangular sub-zone identification. By changing from a single uniform threshold to a multi-parameter spatially-varying confidence index, the system can distinguish between real objects and false positives more effectively, improving both reliability and precision.
Solution Approach 2:
The system provides feedback through the confidence index that reflects the spatial characteristics of each detection. This feedback mechanism allows the driver assistance system to adjust its response based on the calculated confidence level, which incorporates information about detection quality variations across different regions of the sensor's field of view.
3Ease of operation
If distance-based confidence calculation is used, then the calculation method is simple, but detection precision deteriorates due to unrealistic translation of detection quality variation
Solution Approach 1:
The system transitions from a one-dimensional distance-based confidence calculation to a two-dimensional spatial model using triangular sub-zones defined by angular positions and radial distances. This dimensional expansion allows the confidence index to reflect both distance and angular position, providing a more accurate representation of detection quality while maintaining relatively simple geometric calculations.
Data Source
Figure 1~2
Figure 3~4
AI summary
The invention relates to a method for determining a confidence index associated with an object (i) detected by a sensor, comprising steps of approximating the zone of detection by a polygon two sides of which are defined by the lateral borders bounding the angular sector of detection of this sensor, determining the centre of gravity (G) of said polygon, dividing said polygon into a plurality of triangular sub-zones each defined by two adjacent vertices of this polygon and said centre of gravity, identifying, when said sensor detects an object, the triangular sub-zone in which this object is located, determining the equation of the straight line passing through said object and being parallel to the exterior border of this triangular sub-zone, and computing a confidence index associated with said object using a decreasing function (F) of the homothetic ratio (k) with respect to said centre of gravity between said straight line and that defined by said exterior border.