A Three-Camera Digital Image Correlation System For Full-Field 3D Shape and Motion Measurement

Abstract

A cluster-approach-based three-camera digital image correlation (DIC) system is introduced for full-field 3D shape and motion measurement. In this system, three cameras are employed to measure the same specimen area at different viewing angles. Data points within the region of interest can be evaluated by arbitrary camera pairs as a stereo DIC system so that data points with the smallest 3D residuum are selected and mapped into one common coordinate system. Two stationary shape measurements and one out-of-plane motion measurement were carried out with the three-camera DIC system. Test results were analyzed based on the same image series, projection calibration, and correlation parameters, but compared using different camera combinations (i.e., three-camera and two-camera data). Three-camera test results show not only an improved surface coverage due to the additional camera viewing angle for uneven specimen surfaces, but also a smaller and more homogenous distributed measurement uncertainty compared to the two-camera test results. The selection of data points with the smallest 3D residuum evaluated from any arbitrary camera pairs enables a better tolerance of the three-camera DIC system against various measurement error sources such as limited depth of field, lens distortion, and speckle pattern distortion due to tilted camera viewing angles.