Object detection device, object detection method, and industrial vehicle

The article detection device uses an image acquisition and orientation detection system to project ambient images onto a horizontal plane, enabling precise orientation detection of articles from various angles, addressing the limitations of existing technologies.

JP7863827B2Active Publication Date: 2026-05-22TOKYO UNIVERSITY OF SCIENCE +2
View PDF 6 Cites 0 Cited by

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
TOKYO UNIVERSITY OF SCIENCE
Filing Date
2023-02-09
Publication Date
2026-05-22

AI Technical Summary

Technical Problem

Existing article detection devices are limited in their ability to accurately detect the orientation of an article to be handled from a position away from the front direction, necessitating improved methods for precise orientation calculation before the vehicle approaches the article.

Method used

An article detection device that includes an image acquisition unit, an information image creation unit, and an orientation detection unit, which projects ambient images onto a horizontal plane to create an information image with feature lines, allowing for accurate detection of the article's orientation using visual parameter changes and RANSAC for robust feature line detection.

Benefits of technology

Enables accurate detection of the article's orientation regardless of its positional relationship, reducing computational load and enhancing the precision of handling operations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007863827000001
    Figure 0007863827000001
  • Figure 0007863827000002
    Figure 0007863827000002
  • Figure 0007863827000003
    Figure 0007863827000003
Patent Text Reader

Abstract

To provide an article detection device, an article detection method, and an industrial vehicle, which can detect an article to be loaded and unloaded regardless of a positional relationship with the article.SOLUTION: A posture detection section 109 disposes a boundary portion of a feature portion estimated in advance in a pallet 61 at any position of a yaw angle detection information image in the yaw angle detection information image. The posture detection section 109 scans the yaw angle detection information image across a first region E1 in which the feature portion is present and a second region E2 in which the feature portion is not present to acquire a change in a visual parameter. The posture detection section 109 calculates notation information indicating the boundary portion of the pallet 61 on the basis of the change in the visual parameter, and detects a feature line FL on the basis of the notation information. Therefore, in the yaw angle detection information image, the notation information can be written in a portion in which the boundary portion of the pallet 61 is likely to be present. As a result, it is possible to detect the feature line FL accurately.SELECTED DRAWING: Figure 2
Need to check novelty before this filing date? Find Prior Art