Accelerating parallel digital image correlation computation with feature mesh interpolation

Chi, Yuxi; Pan, Bing

doi:10.1016/j.measurement.2022.111554

Cited by 13 publications

(3 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are various IVE methods such as Fast Fourier transform-based cross-correlation (FFT-CC), 8 Optical Flow 9 and feature-based IVE methods like feature-based initial guess (FB-IG), 10 path-independent initial value estimation (PI-IVE). 11 Feature-based IVE is considered to be the most robust and efficient approach in providing the initialization values for large deformation due to its invariance to translation, scale, and rotation.…”

Section: Introductionmentioning

confidence: 99%

“…Moreover, the fewer the detected feature points at a particular region, the higher the probability of mismatched feature points. 9 As a result, more feature points must be computed to achieve densely populated regions of feature points, increasing the computation time. However, several studies have come up with alternatives to deal with the non-uniform distribution of feature points, such as the work conducted by Chi and Pan.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Feature-based template approach for optimizing digital image correlation on complex deformations

Prasad,

Kumar,

Kalra

et al. 2024

Health Monitoring of Structural and Biological Systems XVIII

View full text Add to dashboard Cite

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 featurebased 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.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Feature-based template approach for optimizing digital image correlation on complex deformations

Prasad,

Kumar,

Kalra

et al. 2024

Health Monitoring of Structural and Biological Systems XVIII

View full text Add to dashboard Cite

show abstract

“…Moreover, the local continuity is not considered, leading to potential discontinuities in the resulting dense displacement field. To achieve real-time analysis in local DIC, redundant computations have been eliminated [6] and parallel computing [23] has been employed. Therefore, the calculation of the GLR filed, as a post-processing method, should be calculated as fast as possible to avoid becoming the bottleneck of real-time analysis and to complement the efforts made to accelerate DIC calculation.…”

Section: Introductionmentioning

confidence: 99%

Gray level residual field: an effective metric for pixelwise matching quality evaluation in local digital image correlation

Chi

Zhao

Pan

2023

Meas. Sci. Technol.

Self Cite

View full text Add to dashboard Cite

Gray level residual (GLR), also known as correlation residual field or image difference, reflects the intensity differences between reference and deformed images. In finite element-based global digital image correlation (DIC), GLR field can be readily computed and has been successfully used for crack or damage detection. The computation of pixel-wise GLR field requires dense displacement results, which can be obtained using mesh interpolation in global DIC. To obtain the GLR field in subset-based local DIC, in which results are obtained at discrete calculation points, a simple, efficient and practical computation method for GLR field is presented, which accounts for the influence of illumination variations unavoidably occurred in real DIC tests. The proposed method can be easily implemented on both CPU and GPU to realize real-time visualization. Experiments showed that the GLR field can quantitatively and intuitively evaluate the pixelwise matching quality and provide clearer and finer locations for new components or textures (e.g., cracks) presented in deformed images. By fully leveraging the pixelwise matching quality information provided by the GLR field, the capability and robustness of local DIC is expected to be further enhanced.

show abstract

Improving Deep Learning-Based Digital Image Correlation with Domain Decomposition Method

Chi,

Liu,

Pan

2024

Exp Mech

View full text Add to dashboard Cite

Accelerating parallel digital image correlation computation with feature mesh interpolation

Cited by 13 publications

References 24 publications

Feature-based template approach for optimizing digital image correlation on complex deformations

Feature-based template approach for optimizing digital image correlation on complex deformations

Gray level residual field: an effective metric for pixelwise matching quality evaluation in local digital image correlation

Improving Deep Learning-Based Digital Image Correlation with Domain Decomposition Method

Contact Info

Product

Resources

About