Natural pattern tracking for 3D-digital image correlation measurements

Abstract

Optical methods for structural health monitoring have become extremely popular thanks to advancements in camera technology and increased computational capabilities. Among these methods, 3D-digital image correlation (3D-DIC) and 3D-point tracking (3D-PT) have been proposed as a replacement for traditional contact-based techniques given their ability to measure full-field displacements and surface strain maps of structures of interest. 3D-DIC and 3D-PT require the application of well recognizable high-contrast patterns such as stochastic speckle patterns or optical targets on the surface of the targeted structure. However, when large-scale engineering structures (e.g., bridges and wind turbines) must be analyzed, applying the high-contrast patterns is not always feasible, thus limiting the applicability of optical methods. The objective of this research is to develop an approach that allows identifying and tracking already present features on the surface of the structure of interest, such as bolts, letters, stains, rusted patches, and holes. To achieve this goal, a newly proposed Augmented Centroid-Based Detector (A-CBD) is presented to extract and track identifiable features on the surface of the structure of interest. In this paper, the performance of the A-CBD method is compared with a traditional, correlation-based 3D-PT measurement performed with the use of optical targets in a series of laboratory tests. The results show excellent agreement between the two methods, yielding an average Time Response Assurance Criterion (TRAC) value above 99.5 %. If further developed, the proposed A-CBD approach can extend the applicability of optical methods for measurements on real-world structures.