Vehicle Self-Positioning Using Reflectors and Wavelength Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing techniques for estimating self-position, such as those using feature points from images, face accuracy issues in low-light conditions and high-luminance saturated regions, and can lead to misdetection due to direct light interference from oncoming vehicles.

Innovation Solution

A signal processing system that uses a light projection unit to project light at adjusted wavelengths, detects reflectors, and integrates orientation estimation results from reflector usage, image usage, and GPS/IMU data using Bayesian filtering to enhance self-position estimation accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If feature points are extracted from images to estimate self-position, then the system can recognize surrounding objects and estimate position, but the accuracy is reduced in dark conditions and high-luminance saturated regions

Engineering Contradiction:
Improveself-position estimation accuracyVSAvoidlighting conditions
Core Design Contradiction:
Measurement precisionVSIllumination intensity

Solution Approach 1:

The patent uses a light projection unit to actively change the illumination parameter by projecting light at specific wavelengths onto reflectors in the environment. This transforms the passive imaging condition into an active controlled lighting scenario, ensuring sufficient light intensity for accurate feature extraction regardless of ambient lighting conditions.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces reflectors as intermediary objects that receive projected light and reflect it back to the image pickup device. These reflectors act as mediators between the light source and the imaging system, enhancing the light signal in dark conditions while the saturation judgment unit filters out overly bright saturated regions that would compromise accuracy.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If feature points are extracted from images including high-luminance saturated pixel regions, then the system can process images quickly, but the accuracy of extracting feature points is reduced

Engineering Contradiction:
Improvefeature point extraction accuracyVSAvoidhigh-luminance saturated regions
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent converts the harmful effect of high-luminance saturated regions into a beneficial filtering mechanism. The saturation judgment unit identifies and excludes these saturated pixel regions from feature point extraction, transforming what would be sources of error into a controlled filtering step that improves overall accuracy.

Inventive Principle:
Principle #22Blessing in disguise (Convert harm into benefit)

Solution Approach 2:

The patent extracts and removes the harmful saturated pixel regions from the image processing pipeline. By identifying and excluding these regions before feature point extraction, the system prevents them from degrading the accuracy of position estimation while maintaining efficient processing of valid image data.

Inventive Principle:
Principle #2Taking out (Extraction)

3Adaptability or versatility

If the same light projection system is used by oncoming vehicles, then resource utilization is improved, but direct light from projectors leads to misdetection

Engineering Contradiction:
Improvesystem compatibilityVSAvoiddetection accuracy
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The patent makes the light projection system dynamic by adjusting the wavelength of projected light based on detection results. When another vehicle's projector is detected, the system changes its operating wavelength to avoid interference, transforming a static conflicting scenario into a dynamic adaptive solution that maintains detection accuracy.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent implements feedback control where the detection results from the image pickup device are used to adjust the light projection parameters. When saturation or interference is detected, the system feeds this information back to modify the projection wavelength, creating a closed-loop control system that resolves conflicts with other vehicles' projectors.

Inventive Principle:
Principle #23Feedback

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

The system significantly improves self-position estimation accuracy by reducing interference and maintaining accuracy in low-light conditions and bright environments, while minimizing misdetection from oncoming vehicles.

Implementation Method 1

a light-receiving unit configured to receive light at a predetermined wavelength

Methodology Applied
Scientific EffectLight reception at specific wavelength: Absorption (EM radiation)

Implementation Method 2

a wavelength adjustment unit configured to adjust a wavelength of light to be projected by the light projection unit

Methodology Applied
Scientific EffectWavelength adjustment:

Implementation Method 3

integrates orientation estimation results from reflector usage, image usage, and GPS/IMU data using Bayesian filtering

Methodology Applied
Scientific EffectBayesian filtering:

Data Source

PatentEP3682423B1Signal processing apparatus, signal processing method, program, and moving body
Publication Date: 2021.07.07 SONY GROUP CORP
  • EP3682423B1 patent drawingFigure 1
  • EP3682423B1 patent drawingFigure 2
  • EP3682423B1 patent drawingFigure 3

AI summary

To improve estimation accuracy of a self-position. Light at a predetermined wavelength is projected. An image of a reflector with a reflectance higher than a predetermined reflectance is taken by receiving reflected light of the projected light reflected by the reflector. Own orientation is estimated on the basis of the taken image of the reflector. As a result, the self-position can be highly accurately estimated on the basis of the reflector even at night. The present disclosure can be applied to an on-board system.