Pattern Discrimination Device for Autonomous Train Positioning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current systems for autonomous vehicle navigation, particularly in trains, face challenges in accurately determining the vehicle's position using landmark patterns due to errors in GNSS position information and the need for precise rail shape and gradient data for effective autonomous driving and driver assistance.

Innovation Solution

A pattern discrimination device is developed, equipped with a detector to identify distinct areas in landmark images, a storage unit to correlate landmark patterns with rail information, and a specifying unit to acquire and transmit accurate rail information for position correction, enhancing the accuracy of vehicle positioning.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If GNSS position information is used for vehicle positioning, then positioning can be provided, but position errors occur reducing accuracy

Engineering Contradiction:
Improvepositioning accuracyVSAvoidposition information error
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The system uses landmark patterns as feedback to correct GNSS position errors. The detector identifies landmark patterns in the camera image, compares them with stored patterns, and uses the correlation result to calculate and apply position corrections, forming a closed-loop feedback system that continuously refines positioning accuracy

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

Landmark patterns serve as an intermediary between the vehicle positioning system and the actual position. Instead of relying solely on direct GNSS signals, the system uses landmark patterns as intermediate reference points to indirectly determine and correct the vehicle's position, eliminating direct dependence on potentially erroneous GNSS data

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If landmark patterns are used for position correction, then positioning accuracy improves, but the system complexity increases

Engineering Contradiction:
Improvepositioning accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The landmark pattern database serves multiple functions: it stores reference patterns for identification, provides position correction data, and enables both landmark detection and position calculation. This multi-functionality reduces the need for separate systems and components, thereby limiting the increase in overall system complexity despite the improved positioning accuracy

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Ease of operation

If rail information is correlated with landmark patterns, then autonomous driving support improves, but data processing complexity increases

Engineering Contradiction:
Improveautonomous driving supportVSAvoiddata processing complexity
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

Rail information is pre-correlated with landmark patterns and stored in the database before actual use. When the vehicle needs autonomous driving support, the system simply retrieves the pre-prepared correlated data based on detected landmark patterns, avoiding the need for complex real-time processing and calculations

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4307252A1Pattern discrimination device
Publication Date: 2024.01.17 KK TOSHIBA
  • EP4307252A1 patent drawingFigure 1
  • EP4307252A1 patent drawingFigure 2
  • EP4307252A1 patent drawingFigure 3

AI summary

According to one embodiment, a pattern discrimination device includes a detector configured to detect, in a landmark included in an input image of a front side of a vehicle from an imaging device that is disposed in the vehicle, a first area, and a second area having a higher surface reflection luminance than, or having a different surface reflectance from, the first area; a storage unit configured to store landmark pattern information discriminated from shapes and positions of the first area and the second area, and rail information in a traveling direction of the vehicle, the rail information being correlated with the landmark pattern information; and a specifying unit configured to acquire, from the storage unit, the rail information corresponding to the landmark pattern information corresponding to the first area and the second area detected by the detector.