Geographic Object Detection Using Image and 3D Point Cloud Fusion

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies face challenges in creating highly accurate mapping data for autonomous driving by relying solely on point cloud or image data, as they either fail to specify the three-dimensional position or type of geographic objects accurately, leading to misrecognition or overlooking.

Innovation Solution

A geographic object detection apparatus that combines image data and three-dimensional point cloud data to recognize the type of geographic objects and specify their positions, using a geographic object recognition unit and position specification unit for precise detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If only point cloud data is used for geographic object extraction, then the three-dimensional position can be specified, but the type recognition becomes difficult

Engineering Contradiction:
Improvethree-dimensional position specificationVSAvoidobject type recognition
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent merges point cloud data and image data into a unified detection system. The point cloud data provides three-dimensional position information while image data provides object type recognition. By combining these two data sources through data fusion techniques, the system achieves both accurate position specification and reliable type recognition simultaneously, resolving the contradiction between the two requirements.

Inventive Principle:
Principle #5Merging (Combining)

2Loss of information

If only image data is used for geographic object extraction, then the type can be specified, but the three-dimensional position cannot be accurately determined

Engineering Contradiction:
Improveobject type specificationVSAvoidthree-dimensional position accuracy
Core Design Contradiction:
Loss of informationVSMeasurement precision

Solution Approach 1:

The patent combines image data with point cloud data to overcome the limitations of using only image data. While image data excels at object type specification through visual recognition, the integrated point cloud data supplements this with precise three-dimensional position information. The fusion of these complementary data sources enables simultaneous achievement of both objectives.

Inventive Principle:
Principle #5Merging (Combining)

3Measurement precision

If manual mapping data creation is used, then high accuracy can be achieved, but very high labor costs and temporal costs are required

Engineering Contradiction:
Improvemapping data accuracyVSAvoidlabor cost and time efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent implements an automated mapping data creation system that performs detection and mapping tasks autonomously using integrated point cloud and image data processing. The system automatically identifies geographic objects, determines their positions and types, and generates mapping data without human intervention. This self-service capability maintains high accuracy while dramatically reducing labor costs and processing time compared to manual methods.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces the manual mechanical process of mapping data creation with an automated computational system. Instead of human operators manually analyzing data and creating maps, the system uses algorithmic processing of point cloud and image data to automatically generate mapping information. This substitution of mechanical human labor with automated computational mechanisms achieves both high accuracy and improved productivity.

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

Data Source

PatentEP3876189B1Geographic object detection apparatus, computer-implemented geographic object detection method, and geographic object detection program
Publication Date: 2025.12.24 MITSUBISHI ELECTRIC CORP
  • EP3876189B1 patent drawingFigure 1~2
  • EP3876189B1 patent drawingFigure 3
  • EP3876189B1 patent drawingFigure 4

AI summary

A geographic object recognition unit (120) recognizes, using image data (192) obtained by photographing in a measurement region where a geographic object exists, a type of the geographic object from an image that the image data (192) represents. A position specification unit (130) specifies, using three-dimensional point cloud data (191) indicating a three-dimensional coordinate value of each of a plurality of points in the measurement region, a position of the geographic object.