Road Information Learning Device Curve Registration

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing road information learning systems face challenges in accurately associating curve shapes extracted from map data with those from a vehicle's traveling locus, due to discrepancies and low accuracy in map data, leading to inconsistent curve associations and reduced reliability in driving safety devices.

Innovation Solution

A road information learning device that acquires both map extraction and measurement curves, determines their correspondence through specific conditions, and registers the measurement curve as a definitive registration curve when these conditions are met, enabling accurate road learning aligned with actual travel situations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If map data is used to extract curve shapes for road information, then road information can be obtained, but the accuracy and reliability are reduced due to discrepancies between map data and actual traveling loci

Engineering Contradiction:

Improvecurve shape accuracyVSAvoiddriving safety device reliability

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The patent combines map extraction curves with measurement curves from vehicle traveling loci through a registration process. When correspondence is established between map curves and measurement curves, they are merged to create definitive registration curves that incorporate both the structured map data and the actual vehicle trajectory information, thereby improving both accuracy and reliability

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The patent implements a feedback mechanism where measurement curves from actual vehicle travel are continuously acquired and compared with map extraction curves. The registration results are used to update and refine the road information, creating a closed-loop system that progressively improves accuracy through real-world validation

Inventive Principle:

Principle #23Feedback

2Productivity

If map data is used for road learning, then road information can be acquired, but the association between map curves and measurement curves becomes inconsistent due to map data discrepancies

Engineering Contradiction:

Improveroad information acquisition efficiencyVSAvoidcurve association consistency

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent replaces direct mechanical association between map curves and measurement curves with a computational correspondence determination process. Instead of relying on geometric matching alone, the system uses a multi-step process involving curve extraction, correspondence determination based on multiple criteria, and conditional registration to establish associations, thereby improving consistency

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the parameters used for curve association from simple geometric matching to a comprehensive set of correspondence criteria including positional relationships, curvature characteristics, and contextual road features. This parameter transformation enables more reliable and consistent association between map curves and measurement curves

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11287265B2Road information learning device

Publication Date: 2022.03.29 DENSO CORP

AI summary

A road information learning device acquires a map extraction curve and a measurement curve for a traveling curve along which a vehicle travels, determines whether the map extraction curve and the measurement curve correspond to an identical traveling curve, and registers the measurement curve as a definitive registration curve when the map extraction curve and the measurement curve correspond to the identical traveling curve, and a predetermined definitive registration condition is satisfied.