Machine vision-based techniques for non-contact structural health monitoring

Inventors

Wu, Zheng Yi • Mo, Dian • Xiao, Peng

Assignees

Bentley Systems Inc

Abstract

In an example embodiment, a structural health monitoring software application provides non-contact structural health monitoring using a video of a structure captured by a video camera. The application selects an area of interest and divides the area of interest into a grid of cells. One or more machine vision algorithms are selected from a set of multiple machine vision algorithms provided by the application, wherein the set of multiple machine vision algorithms includes at least one phase-based algorithm and at least one template matching algorithm. The application applies the one or more machine vision algorithms to the video of the structure to determining a displacement of each cell, detects defects or damage based on differences in the displacement of cells, and displays an indicator of the detected defects or damage in a user interface.

Core Innovation

The invention relates to non-contact structural health monitoring using a structural health monitoring software application executing on one or more computing devices to obtain a video of a structure captured by a video camera. The software selects an area of interest in frames of the video and divides the area of interest into a grid of cells. It selects machine vision algorithms to determine a displacement of each cell, and uses the resulting per-cell displacement information for defect or damage detection.

The system determines displacement for each cell using machine vision algorithms including at least one phase-based algorithm and at least one template matching algorithm. For template matching, the method defines templates in an initial frame, searches for each template in subsequent frames by sliding the template, comparing the template to an image patch at possible locations, generating a similarity map, and determining a best-matching image patch to calculate displacement based on template and best-match locations across frames.

For phase-based processing, the method selects interested points in an initial frame, selects an averaging area for each interested point, calculates local phases over subsequent frames, and tracks phase change for the interested points within respective neighborhoods. After displacement determination, the method detects defects or damage based on differences in the displacement of cells, and displays an indicator of the detected defects or damage in a user interface on a screen of the computing devices.

Claims Coverage

The independent claims are directed to two method variants and computing device implementations. Across these independent claims, the main inventive coverage contains four inventive features: non-contact video-based SHM with ROI selection and grid-of-cells partitioning, displacement determination using selected machine vision algorithms, defect or damage detection based on differences in per-cell displacement directions, and displaying an indicator of detected defects or damage in a user interface on a screen.

Overall, the independent claims cover a non-contact SHM software workflow that partitions a video-selected region into a grid, computes per-cell displacement using template matching or phase-based local phase tracking, detects defects or damage from differences in displacement directions, and displays an indicator of detected defects or damage in a user interface; some independent claim variants support comparative indicator outputs derived from phase-based and template matching approaches.

Stated Advantages

Documented Applications