Markers are crucial for measuring the displacement of large structures using digital image correlation (DIC). In general, creating or affixing artificial markers on large structures is challenging. Several studies in the past have addressed this challenge by employing marker-less or target-free approaches. In this work, we integrate an AI-based non-intersecting poly-shape marker identification technique with the DIC program that uses the structural pattern as a marker and automates the selection of the region of interest by enabling strong correlation criteria to obtain the displacements in real-time. The proposed AI-DIC algorithm segments non-intersecting poly-shape markers from the images of the target structure based on the features detected by the KAZE feature detector and descriptors. Further, the investigated marker or the natural structural pattern is automatically given as an input region of interest to the DIC program. Moreover, it considers the marker as a template, correlates it with all subsequent images, and analyzes the displacements and frequencies of the target structure. In addition, the AI-DIC algorithm is realized on an in-house cantilever beam experiment where the images are acquired and processed in real-time.
The initialization method in Digital Image Correlation (DIC) is essential for optimizing the correlation criteria and accurately computing the deformations of a material under load. At present, feature-based initialization techniques are widely explored for predicting the deformations of various complex circumstances, such as large deformations for soft materials, non-continuous deformations in heterogeneous materials, etc. However, due to the non-uniform distribution of the detected features, the initialization process goes through biased prediction. This bias occurs due to the sparsity of features in different regions of the sample, which can lead to inaccuracy in identifying the shape of deformation. This study addresses the issue of feature distribution and develops a feature-based template approach for providing initialization points for each subset on a finer scale. The features (interest points) are determined using KAZE feature detector and descriptor algorithm in nonlinear scale space due to its ability to determine consistent, repeatable, distinct features invariant to scale and rotation. The proposed algorithm uses bi-cubic b-spline interpolation to identify the strongest interest point at the subpixel level for each subset (of the input sample images), which works as an initial value for estimating the deformation. Further, a threshold-based incremental reference approach is developed for measuring large deformations and avoiding the cumulative errors associated with the commonly used incremental reference strategy, which is compute-intensive because of the comparison between every previous image and the subsequent images.
An automatic lightweight feature detection algorithm is developed to perform real-time structural health monitoring (SHM) of large structures. The algorithm works on the specified region of interest (ROI) and applies canny edge detection with k-means clustering for identifying the displaced pixel location in an image sequence. The location of detected edges (white pixels) in the selected ROI is first validated and then given as input to the k-means clustering algorithm for centroid calculation. The pixel movement tracing method is validated by image simulation, indoor digital micrometer experiment and then an outdoor field experiment on wind turbine. The image simulation experiment was performed to generate sample data and ground truth values. In this experiment, the algorithm was able to detect the defined pixel translations. With this validation, other two experiments were conducted. The indoor experiment was implemented for experimental verification where it successfully identifies the moving bar’s 20mm displacement. Likewise, it also accurately measures the natural frequency of the tower of a utility-scale wind turbine. Hence, the algorithm was built on parallel processing with multi-ROI selection to optimize the space and time complexity for real-time vibration analysis. The present study proclaims that the developed algorithm can be used to perform real-time SHM of large-scale structures.
While developing a novel digital image correlation (DIC)-based NDT method, one has to integrate an automatic and robust feature detection method with the DIC technique. Several studies in the past employed various algorithms such as SIFT, SURF and BRISK with DIC for feature detection and correlation initiation purposes. However, our study shows that the performance of available algorithms is subjected to the image of interest from a particular field experiment. Therefore, the selection of the feature detection algorithms is an essential step towards accurate and efficient processing. We have developed a methodology that applies various feature detection algorithms (namely SIFT, SURF, BRISK, ORB and KAZE) and selects the most accurate, efficient and repeatable algorithm for detecting unique natural patterns. Moreover, the methodology is integrated with an in-house 3D-DIC program to identify as well as correlate natural patterns to obtain in-plane and out-of-plane displacements of large structures. The combined methodology is successfully applied and verified by performing field experiments with a light tower of 10m height and a utility-scale wind turbine. It is observed that the developed methodology is robust enough to detect natural patterns accurately and efficiently. It has also been demonstrated that the technique is successful with the determination of 3D displacements and natural frequencies of the large structures.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.