A new approach for 2-D and 3-D precise measurements of ground deformation from optimized registration and correlation of optical images and ICA-based filtering of image geometry artifacts

Fault Friction Derived From Fault Bend Influence on Coseismic Slip During the 2019 Ridgecrest Mw 7.1 Mainshock

“…To measure the tectonic surface deformation pattern we use a new optical image correlation technique that we have developed called COSI‐Corr + (Aati et al., 2022). We apply this open‐source and automated image processing technique for the first time to resolve the full 3D deformation field of the Ridgecrest earthquake sequence as this method offers a number of benefits over current image matching approaches.…”

“…The general COSI‐Corr + workflow involves five main steps, this includes (a) the RSM refinement, (b) image orthorectification and resampling, (c) sub‐pixel image correlation, (d) 3D displacement calculation via ray tracing and (e) deconstruction of the 3D deformation maps with ICA (for additional details see Aati et al. [2022]). This workflow results in a final set of three deformation maps where the surface motion is decomposed into the east‐west, north‐south and vertical component of motion (see Figure S1 in Supporting Information S1).…”

“…Here we find the uncertainty in the east-west, north-south, and vertical directions is 0.7, 0.6 and 0.6 m at the 90% confidence level, respectively. This processing workflow results in a significant reduction of topographic, orthorectification, and CCD array artifacts (with a reduction of uncertainty in the deformation maps by a factor of ∼3.6 compared to a result using the traditional "2 + 1 D approach"), and near complete removal of aliasing artifacts associated with the SPOT 6 images (see Aati et al [2022] for details).…”

“…The latter is especially useful as it provides a greater number of satellite look vectors with which to more accurately triangulate the 3D position of pixels. Additional processing details are described in Aati et al (2022).…”

Identification of bridge modal parameters from GNSS data by integrating IEWT and robust ICA algorithm

“…Using the higher-order statistics of data, ICA divides a signal into linear combinations of statistically independent non-Gaussian signals [18]. ICA has been widely used in image processing, speech signal processing, biomedical signal processing, pattern recognition, and SHM, due to its superiority in blind source separation [19][20][21]. Barbosh et al [22] combined multivariate empirical mode decomposition algorithm with ICA to.…”