Computer vision-based system and method for assessment of load distribution, load rating, and vibration serviceability of structures

Patent metadata

Field	Value
Publication number	US-12430733-B2
Application number	US-202217712038-A
Country	US
Kind code	B2
Filing date	Apr 1, 2022
Priority date	Jun 1, 2021
Publication date	Sep 30, 2025
Grant date	Sep 30, 2025

Abstract

A computer vision-based system provides for load distribution estimation and load rating and vibration serviceability assessment of structures. The system integrates evaluates the structural load carrying capacity, the diagnosis and prognosis of performance and safety, and vibration serviceability. Cameras record images of a structure, and regions of interest are monitored in those images for their displacement and velocity as loading varies. Where the displacement determined exceeds a predetermined threshold, or where the acceleration determined exceeds predetermined limits, or where the distribution of displacements of parts of the structure deviates substantially from an estimated displacement distribution, an output indicating potential problems with the structure is output.

First claim

What is claimed is:

1. A computer-vision-based method for determining load worthiness, safety, and/or serviceability of a structure, said method comprising:
directing one or more cameras at the structure so as to derive therefrom image data defining a series of images of target portions of the structure;
receiving the image data at a computer system;
applying a computer vision process performed by said computer system to said image data so as to derive location data from the image data, the computer vision process being a feature-extraction process in which location data from the target portions are derived from key points in the images such that said location data corresponds to locations of the target portion in each of said images; determining displacement data from the location data for the target portions by comparison of the location data for each of the target portions with respective original location data for each of the target portions; determining a safety and/or serviceability assessment for said structure from said displacement data, wherein determining of the safety and/or serviceability assessment for said structure includes comparison of the displacements of each of the target portions to an estimated displacement distribution, and determination of a structural problem for said assessment responsive to a determination that the displacement of one or more of the target portions deviates from the estimated displacement distribution by more than a predetermined limit; and outputting a report of the safety and/or serviceability assessment.

2. The computer-vision-based method of claim 1 , wherein the determining of the safety and/or serviceability assessment for said structure includes deriving from the location data velocity data and acceleration data indicative of, respectively, a real-world velocity and a real-world acceleration of the of the target portion.

3. The computer-vision-based method of claim 2 , wherein the safety and/or serviceability determination includes comparing said displacement with a predetermined maximum displacement for said target portion, and
wherein said safety and/or serviceability assessment includes an indication of a potential problem with the structure responsive to a determination that the displacement in greater than said predetermined maximum displacement.

4. The computer-vision-based method of claim 1 , wherein the structure is a bridge having a plurality of longitudinally extending, laterally spaced girders, said target portion being at a longitudinal distance along one of the girders, and
said camera or another camera or cameras providing image data defining a series of images of other target portions of the structure that are each located on a respective one of the other girders at the same longitudinal distance therealong such that the target portions are generally aligned laterally; and
said applying of the computer vision process performed by said computer system derives location data for the other target portions from the image data corresponding to locations of the target portion in each of said images, said location data being indicative of a deflection of the respective girder.

5. The computer-vision-based method of claim 1 , wherein the method further comprises calibrating the computer system such that the computer system has stored accessible thereto calibration data providing for conversion of distances of pixels in the images to real-world distance units.

6. The computer-vision-based method of claim 5 , wherein the images are digital data defining pixels of the images, and the location data is derived from a location of the target portion as a pixel location thereof in the image, said pixel location being converted to real-world units using the calibration data.

7. The computer-vision-based method of claim 1 , wherein the computer vision process includes applying computer vision to locate the target portion in the image by searching only within a predetermined searching area smaller than the image.

8. The computer-vision-based method of claim 1 , wherein the target portion is marked with a pattern that makes said target portion readily visible to the computer vision process.

9. The computer-vision-based method of claim 1 , wherein the target portion is a structural component of the structure that is selected due to visibility thereof to the computer vision process.

10. The computer-vision-based method of claim 1 , wherein the system has at least two of said cameras each producing a respective series of images of respective target portions of the structure at a sequence of equally spaced points in time, and the computer vision process comprises
performing a comparison, for each of the target portions, of a pixel location of the target portion in a first of the images with pixel locations of said target portion in the images later in the series of images and deriving for each of the target portions from said comparison one or more displacement data values as a number of pixels in the associated image.

11. The computer-vision-based method of claim 10 , wherein the method further comprises the computer system
performing a comparison, for each of the target portions, of the pixel location of the target portion in each of the images with the pixel locations of said target portion in the next of said images in the series of images and deriving from said comparison a series of velocity data values as a number of pixels in the associated image; and
converting the velocity data to real-world measurement units using calibration data stored by the computer for each of the cameras defining a scaling of the pixels thereof to real-world measurements of movement of the respective target portion; and
calculating a derivative of said velocity data so as to derive therefrom acceleration data corresponding to acceleration of the target portion in real-world measurements; and
wherein the determining of the safety and/or serviceability assessment for said structure includes comparison for each of the target portions of the acceleration indicated by the acceleration data with a predetermined acceleration limit, and determination of a structural problem responsive to the comparison indicating one or more of the accelerations being greater than the predetermined acceleration limit.

12. The computer-vision-based method of claim 10 , wherein the computer has data indicative of the estimated displacement distribution of the target portions subject to loading of the structure during the period of taking of the images of the target portions.

13. The computer-vision-based method of claim 12 , wherein the structure is a bridge, and the method includes
receiving image data from a further camera positioned corresponding to video of a portion of the bridge on which vehicles travel, and
wherein the estimated displacement distribution is selected based on a detected position of a vehicle or positions of vehicles on the structure.

14. A computer-vision-based method for determining load worthiness, safety or serviceability of a structure, said method comprising:
directing one or more cameras at the structure so as to derive therefrom image data defining a series of images of a target portion of the structure;
receiving the image data at a computer system;
applying a computer vision process performed by said computer system to said image data so as to derive location data from the image data, said location data corresponding to locations of the target portion in each of said images;
determining a safety and/or serviceability assessment for said structure from said location data; and
outputting a report of

Assignees

• Univ Central Florida Res Found Inc
• Univ Of Central Florida Research Foundation

Inventors

• Catbas Fikret Necati
• Dong Chuanzhi

Classifications

• H04N23/90

  Arrangement of cameras or camera modules, e.g. multiple cameras in TV studios or sports stadiums · CPC title

• G06T7/74

  involving reference images or patches · CPC title

• G06V20/54

  of traffic, e.g. cars on the road, trains or boats · CPC title

• G06T2207/30184

  Infrastructure · CPC title

• G06V10/225 Primary

  based on a marking or identifier characterising the area · CPC title