Lane Line Identification via Vanishing Point and Trajectory Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current lane line identification methods, such as manual identification and trained models, are inefficient and prone to errors, especially in scenarios not included in the training data, and often misidentify road features like railings or streetlights as lane lines.

Innovation Solution

An automated system that uses image acquisition devices to capture road images, identifies reference straight lines, determines a vanishing point, and selects candidate lane lines based on this point, along with travel trajectories to accurately identify lane lines, applicable to various scenarios.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual lane line identification is used, then accuracy can be maintained through human judgment, but efficiency deteriorates due to time-consuming processes

Engineering Contradiction:
Improvelane line identification accuracyVSAvoididentification efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system performs automated lane line identification without human intervention. The processing device automatically captures images, detects straight lines, determines vanishing points, and identifies lane lines through algorithmic operations, making the system self-sufficient and eliminating manual labor while maintaining high efficiency

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical identification with an automated image processing system. The processing device uses computer vision algorithms including straight line detection, vanishing point calculation, and lane line identification algorithms to substitute human visual inspection and manual marking operations

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If trained lane line identification models are used, then processing speed improves, but accuracy deteriorates in scenarios not included in training data

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidlane line identification accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system dynamically adjusts identification parameters based on scene characteristics. By calculating the vanishing point from detected straight lines and using it to guide lane line identification, the system adapts to different road scenarios (highways, crossroads, etc.) without requiring retraining, maintaining accuracy across diverse conditions

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The vanishing point-based identification method serves as a universal solution for multiple road scenarios. The same algorithmic approach works for highways, crossroads, and other road types by adapting to the geometric characteristics of each scene through vanishing point calculation, eliminating the need for scenario-specific training models

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

3Productivity

If line detection algorithms are used, then processing speed improves, but accuracy deteriorates due to misidentification of road features

Engineering Contradiction:
Improveprocessing efficiencyVSAvoidlane line identification accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The vanishing point serves as an intermediary element that connects straight line detection to lane line identification. By using the vanishing point as a reference, the system can distinguish lane lines from other straight road features (railings, streetlights) because lane lines converge at the vanishing point while other features do not, thereby improving accuracy without sacrificing efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3821371B1Methods and systems for lane line identification
Publication Date: 2023.05.24 ZHEJIANG DAHUA TECH CO LTD
  • EP3821371B1 patent drawingFigure 1
  • EP3821371B1 patent drawingFigure 2
  • EP3821371B1 patent drawingFigure 3

AI summary

The present disclosure provides a method for identifying a plurality of lane lines on a road segment. The method may include obtaining a plurality of images of the road segment having one or more lanes and vehicles, and identifying a plurality of reference straight lines from the images. The method may include determining a vanishing point based on the reference straight lines, and determining a plurality of candidate lane lines among the reference straight lines based on the vanishing point. The method may also include determining at least one travel trajectory of at least one of the vehicles based on the images. The method may further include determining a target travel trajectory among the travel trajectory based on the vanishing point or a count of the lanes, and determining the lane lines among the candidate lane lines based on the target travel trajectory.