Detecting Degraded Ground Paint for Accurate Virtual Control Points

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for determining ground control points from images of road markings face challenges due to degradation of painted road markings, which affects the accuracy of digital map data required for location-based services like autonomous driving.

Innovation Solution

A computer-implemented method and apparatus for detecting degraded ground paint in images using semantic segmentation, binary image generation, and hole-filling techniques to identify and correct faded or damaged road markings, enabling the generation of accurate virtual ground control points.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If painted road markings are used as ground control points, then the cost and time for map data collection is reduced, but the accuracy of ground control points deteriorates due to paint degradation

Engineering Contradiction:
Improvemap data collection efficiencyVSAvoidground control point accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary detection of paint degradation using image processing and semantic segmentation before the paint markings become completely unusable. By identifying degraded areas in advance, the system can either correct them through image processing or flag them for replacement, ensuring that ground control points maintain their accuracy throughout the mapping process

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediate detection and classification layer between the painted road markings and the ground control point identification process. This intermediary system uses semantic segmentation and image analysis to assess the quality of paint markings, filtering out degraded markings before they compromise the accuracy of ground control points while allowing healthy markings to be used efficiently

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If traditional ground control points are used for precise alignment, then the accuracy of digital map data is improved, but the cost and complexity of data collection increases

Engineering Contradiction:
Improvedigital map data accuracyVSAvoiddata collection system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system creates a digital copy or representation of ground control points by using painted road markings captured in images. Instead of requiring physical ground control points, the system processes images to identify and validate virtual ground control points based on paint markings, thereby simplifying the data collection process while maintaining accuracy through image-based verification

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces the mechanical/physical system of traditional ground control points with an image processing-based system. By using semantic segmentation and computer vision algorithms to identify and validate ground control points from painted road markings in images, the system eliminates the need for physical measurement equipment and manual ground control point establishment, reducing complexity while preserving accuracy

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

3Speed

If degraded paint markings are not detected, then the processing speed is maintained, but the reliability of ground control points deteriorates

Engineering Contradiction:
Improveprocessing speedVSAvoidground control point reliability
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The system applies partial action by selectively processing only those image regions containing painted road markings, rather than analyzing entire images. The semantic segmentation focuses computational resources on relevant areas, and degradation detection is performed only on identified paint regions, maintaining processing speed while ensuring reliability through targeted analysis of critical areas

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3706072B1Method, apparatus, and system for detecting degraded ground paint in an image
Publication Date: 2022.08.10 HERE GLOBAL BV
  • EP3706072B1 patent drawingFigure 1
  • EP3706072B1 patent drawingFigure 2
  • EP3706072B1 patent drawingFigure 3

AI summary

An approach is provided for detecting degraded ground paint in an image. The approach, for example, involves performing semantic segmentation on the image to determine one or more pixels of the image that are classified in a ground paint category. The approach also involves generating a binary image that contains the one or more pixels of the image that are classified in the ground paint category. The approach further involves generating a hole-filled binary image by filling in the binary image to generate one or more curvilinear structures from the one or more pixels. The approach further involves determining a difference between the image and the hole-filled binary image to identify one or more degraded ground paint pixels of the image and providing the one or more degraded ground paint pixels as an output.