Camera-Based HD Map Creation Using Feature Point Tracking

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The existing high-definition map creation and updating process for autonomous vehicles is costly and inefficient, relying on expensive mobile mapping systems that struggle to quickly adapt to road changes, which can compromise safety.

Innovation Solution

A camera-based system that uses Ground Control Points (GCPs) and feature points to recognize and create high-definition maps, reducing the need for expensive mobile mapping systems by employing camera-equipped probe vehicles to update maps in real-time, thereby decreasing communication loads and costs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If mobile mapping systems (MMS) are used to create and update high-definition maps, then map accuracy and detail are improved, but cost and labor requirements increase significantly

Engineering Contradiction:
Improvemap accuracyVSAvoidcost
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent uses standard cameras to capture images that are processed into point cloud data, creating a simplified copy of the complex MMS data collection process. This allows map creation without expensive specialized equipment while maintaining sufficient accuracy for autonomous driving applications.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The invention replaces expensive, complex MMS equipment with inexpensive standard cameras that can be easily deployed on multiple vehicles. These cameras capture sufficient data for map creation without requiring the high cost of professional mapping systems.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

2Quantity of substance

If mobile mapping systems gather extensive data per hour, then map detail is improved, but real-time processing capability deteriorates

Engineering Contradiction:
Improvedata volumeVSAvoidupdate speed
Core Design Contradiction:
Quantity of substanceVSProductivity

Solution Approach 1:

The patent extracts only the essential features from camera images to create point cloud data, rather than processing complete high-resolution images. This selective extraction reduces data volume significantly while maintaining the key information needed for autonomous driving navigation.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system performs partial processing by focusing on extracting critical spatial features and geometric information from images, rather than processing all image data. This partial action approach enables faster processing speeds suitable for real-time map updates.

Inventive Principle:
Principle #16Partial or excessive action

3Reliability

If multiple probe vehicles are deployed to update maps in real-time, then map freshness is improved, but communication load and costs increase

Engineering Contradiction:
Improvemap freshnessVSAvoidcommunication cost
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

The patent creates compact representations of map data from camera images that require less communication bandwidth. By processing images locally into point cloud data and essential map features, the system reduces the communication load when transmitting map updates to central servers or other vehicles.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3842751B1System and method of generating high-definition map based on camera
Publication Date: 2024.02.21 U1GIS
  • EP3842751B1 patent drawingFigure 1
  • EP3842751B1 patent drawingFigure 2~3
  • EP3842751B1 patent drawingFigure 4~5

AI summary

According to an embodiment, there is provided a system creating a high-definition map based on a camera (120). The system includes at least one map creating device (100) that includes an object recognizing unit (111) recognizing, per frame of the road image, a road facility object including at least one of a GCP object and an ordinary object and a property, a feature point extracting unit (112) extracting a feature point of at least one or more road facility objects from the road image, a feature point tracking unit (113) matching and tracking the feature point in consecutive frames of the road image, a coordinate determining unit (115) obtaining relative spatial coordinates of the feature point to minimize a difference between camera pose information predicted from the tracked feature point and calculated camera pose information, and a correcting unit (116) obtaining absolute spatial coordinates of the feature point.