Parallax Detection in Multi-Sensor Data via Sorting Order Comparison

Solution Overview

Problem

Modern transportation vehicles face challenges in accurately associating sensor data from different sensors due to the parallax problem, which can lead to incorrect distance estimates during sensor data fusion, especially when sensors are positioned differently and capture the environment from different directions.

Innovation Solution

A method and apparatus that detect parallax problems by assigning measured values from one sensor to corresponding values from another, taking into account the respective imaging conditions, and checking the sorting order of these values to determine if a parallax issue exists, using an input device to receive sensor data, a computing device to perform the assignment and check for matching sorting orders, and an output device to verify the result.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If sensors are arranged at different positions to capture the environment from different directions, then the coverage and information collection capability are improved, but a parallax problem occurs leading to incorrect association of sensor data

Engineering Contradiction:

Improveenvironment capture capabilityVSAvoiddistance estimate accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent replaces complex geometric and physical calculations for resolving parallax with a simple sorting order comparison method. Instead of using mechanical or mathematical models to compute parallax corrections, the system uses a logical sorting check that compares the sequential order of measured values from different sensors, significantly simplifying the problem while maintaining accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the approach from comparing actual measured values to comparing the sorting order of measured values. This parameter transformation converts a complex spatial association problem into a simple ordinal comparison problem, where the system checks whether the sequence of values from one sensor matches the sequence from another sensor, indicating the presence or absence of parallax effects.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If complex parallax correction algorithms are used to improve distance estimation accuracy, then measurement precision is improved, but device complexity and computing power requirements increase

Engineering Contradiction:

Improvedistance estimation accuracyVSAvoidcomputing device complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex geometric and physical calculations for resolving parallax with a simple sorting order comparison method. Instead of using mechanical or mathematical models to compute parallax corrections, the system uses a logical sorting check that compares the sequential order of measured values from different sensors, significantly simplifying the problem while maintaining accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent employs a computationally inexpensive method that requires minimal processing power. The sorting order comparison uses simple logical operations rather than intensive mathematical computations, making the system efficient and suitable for implementation in resource-constrained environments such as automotive assistance systems.

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

3Area of stationary object

If sensors are positioned at different locations, then the sensing coverage is improved, but the association of corresponding measured values becomes difficult leading to data fusion problems

Engineering Contradiction:

Improvesensing coverage areaVSAvoidmeasured value association difficulty

Core Design Contradiction:

Area of stationary objectVSDifficulty of detecting and measuring

Solution Approach 1:

The patent changes the approach from comparing actual measured values to comparing the sorting order of measured values. This parameter transformation converts a complex spatial association problem into a simple ordinal comparison problem, where the system checks whether the sequence of values from one sensor matches the sequence from another sensor, indicating the presence or absence of parallax effects.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces complex geometric and physical calculations for resolving parallax with a simple sorting order comparison method. Instead of using mechanical or mathematical models to compute parallax corrections, the system uses a logical sorting check that compares the sequential order of measured values from different sensors, significantly simplifying the problem while maintaining accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS11704824B2Method and device for determining a parallax problem in sensor data of two sensors

Publication Date: 2023.07.18 VOLKSWAGEN AG

AI summary

A method for detecting a parallax problem in sensor data from two sensors spaced apart from each other and at least partly capturing the same environment, wherein at least one of the sensors provides distance information. The method includes obtaining the acquired sensor data from the sensors; assigning measured values in the acquired sensor data of one sensor to corresponding measured values in the acquired sensor data of the other sensor, wherein the assignment takes the respective imaging conditions of the two sensors into account; consecutively numbering the measured values in the sensor data; checking whether a sorting order of the numbering of the measured values that correspond to each other matches, a parallax problem being determined in response to a sorting order not matching; and outputting a test result. Also disclosed is an apparatus for detecting a parallax problem in sensor data from two sensors and a transportation vehicle.