A bridge intelligent patrol disease position identification and three-dimensional visualization marking method

By constructing a BIM model and a defect identification AI model, and combining the relationship between pixel displacement and viewpoint change, the automatic location and visual identification of bridge defects in three-dimensional space were realized, which solved the problem of inaccurate defect coordinate acquisition in existing technologies and improved inspection efficiency and accuracy.

CN122244140APending Publication Date: 2026-06-19广东建科创新技术研究院有限公司

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
广东建科创新技术研究院有限公司
Filing Date
2026-03-26
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing technologies cannot automatically obtain the precise coordinates of bridge defects in three-dimensional space, resulting in low bridge inspection efficiency and limited coverage, making it difficult to meet the requirements of refined management in modern bridge operation and maintenance.

Method used

By constructing a BIM model of multi-source bridge data, collecting image metadata and inputting it into the disease identification AI model, identifying the pixel coordinates of the disease area, and combining the relationship between pixel displacement and viewing angle change, adjusting the viewing angle of the shooting equipment, the disease can be automatically located and visualized in three-dimensional space.

Benefits of technology

It enables high-precision automatic positioning and visualization of defects in three-dimensional space, reduces the complexity of equipment deployment, improves inspection efficiency and accuracy, and enhances the digitalization level of bridge operation and maintenance management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122244140A_ABST
    Figure CN122244140A_ABST
Patent Text Reader

Abstract

This invention relates to the field of bridge inspection technology and discloses a method for intelligent bridge inspection, including the identification and 3D visualization of bridge defects. The method comprises: acquiring multi-source bridge data and constructing a BIM model; collecting bridge surface images and extracting image metadata; inputting the image metadata into a defect identification AI model to determine the pixel difference between the pixel coordinates of the defect area and the pixel coordinates of the center point; determining the viewing angle adjustment amount of the shooting device based on the relationship between pixel displacement and viewing angle change, shooting device information, and pixel difference; obtaining the initial deviation parameters between the shooting device and the virtual viewpoint in the BIM model; determining the target viewing angle parameters of the BIM model; adjusting the virtual viewpoint to the target viewing angle parameters; and generating defect identification objects in 3D spatial coordinates. This invention achieves the mapping from defect pixel coordinates in bridge surface images to 3D spatial coordinates, improving the spatial positioning accuracy of defects.
Need to check novelty before this filing date? Find Prior Art