3D Point Cloud Target Identification for Work Vehicles

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing target identification systems face challenges in accurately identifying targets using image capturing conditions and three-dimensional data, particularly due to brightness issues and high processing loads.

Innovation Solution

An identification system and method that utilizes a three-dimensional sensor to detect objects, estimates a plane from point cloud data, extracts point clouds on the plane to generate a binary image, and identifies targets from this image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If target identification is performed using three-dimensional data from a three-dimensional sensor, then identification accuracy is improved, but processing load increases and real-time calculation becomes difficult

Engineering Contradiction:
Improveidentification accuracyVSAvoidprocessing load
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The point cloud data is segmented into multiple regions based on depth information and spatial distribution. By dividing the three-dimensional space into foreground, background, and intermediate regions, the system processes only relevant portions of the data in detail while simplifying distant regions, thereby reducing overall processing load while maintaining identification accuracy for targets of interest

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system extracts only the essential features and characteristics from the complete three-dimensional point cloud data. By identifying and extracting key geometric features, depth information, and spatial relationships necessary for target identification, the system reduces the data volume requiring processing while preserving the information needed for accurate identification

Inventive Principle:
Principle #2Taking out (Extraction)

2Device complexity

If target identification is performed using images, then processing load is reduced, but identification accuracy deteriorates due to capturing conditions such as brightness

Engineering Contradiction:
Improveprocessing loadVSAvoididentification accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The system transitions from two-dimensional image data to three-dimensional point cloud data for target identification. By utilizing depth information and spatial coordinates from the three-dimensional sensor, the system overcomes the limitations of image-based methods that are sensitive to brightness and lighting conditions, achieving more robust and accurate identification while managing processing load through selective processing strategies

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP4730257A1Identification system, identification method, and work vehicle
Publication Date: 2026.04.22 KOMATSU LTD
  • EP4730257A1 patent drawingFigure 1
  • EP4730257A1 patent drawingFigure 2~3
  • EP4730257A1 patent drawingFigure 4~5

AI summary

An identification system 1 includes a three-dimensional sensor 11 that is attached to a work vehicle and detects objects present around the work vehicle, and a controller 20. The controller 20 acquires point cloud data that is object detection data from the three-dimensional sensor 11, estimates a plane constituting at least a part of an outer shape of a target from the point cloud data, extracts a point cloud located on the estimated plane to generate a binary image, and identifies the target from the binary image.