DICpy.DIC_2D

The module DIC_2D of DICpy contain classes for performing 2D digital image correlation with Python. The module follows the objected-oriented paradigm and the DIC is performed following a sequence of object instantiation. Next, a detailed description of each class is presented.

DICpy.DIC_2D.Images

class DICpy.DIC_2D.Images[source]

Bases: object

This class contains the methods for reading and processing images used in digital image correlation (DIC). Input: Attributes:

  • path_speckle (str)

    Path to the speckle images.

  • path_calibration (str)

    Path to the calibration image.

  • ref_image (ndarray)

    Reference image.

  • calibration_image (ndarray)

    Calibration image.

  • images (ndarray)

    Gray images.

  • images_normalized (ndarray)

    Gray images normalized by 255.

  • lx (int)

    Number of columns of each image.

  • ly (int)

    Number of rows of each image.

  • num_img (int)

    Number of images.

  • pixel_dim (float)

    Size of each pixel in length dimension.

Methods:

calibration(ref_length=None, pixel_dim=None, path=None, file_name=None, point_a=None, point_b=None, verbose=True)[source]

Method for the analysis calibration.

Input: * ref_length (str)

Length used as reference.

  • pixel_dim (float)

    Size of each pixel in length dimension.

  • path (str)

    Path for the calibration image used in the DIC analysis.

  • file_name (str)

    Name of the calibration image.

  • point_a (float)

    First corner of the region of interest (ROI).

  • point_b (float)

    Corner opposed to point_b of the region of interest (ROI).

  • verbose (bool)

    Boolean variable to print some information on screen.

Output/Returns:

read_speckle_images(path=None, extension=None, file_names=None, ref_id=0, verbose=False)[source]

Read the speckle images.

Input: * path (str)

Path for the images used in the DIC analysis.

  • extension (str)

    The extension of the files located in path.

  • file_names (str)

    Define the files to be read in the sequence of your preference.

  • ref_id (int)

    Define a file to be used as reference, the default is zero, which means that the first image in the stack will be used as reference.

  • verbose (bool)

    Boolean varible to print some information on screen.

Output/Returns:

DICpy.DIC_2D.RegularGrid

class DICpy.DIC_2D.RegularGrid(images_obj=None)[source]

Bases: object

This class contains the methods for creating a rectangular mesh used in the DIC analysis. Input: * images_obj (object)

Object containing the speckle images, reference images, and calibration images, as well as calibration parameters.

Attributes:

  • images_obj (object)

    Object containing the speckle images, reference images, and calibration images, as well as calibration parameters.

  • stepx (ndarray)

    Steps for the grid in the x direction (columns).

  • stepy (ndarray)

    Steps for the grid in the y direction (rows).

  • centers (list)

    Centers for each cell in the grid.

  • wind (list)

    Dimensions in both x and y dimensions for each cell.

  • xp (ndarray)

    Position of the upper right corner for each cell (grid in the x direction - columns).

  • yp (ndarray)

    Position of the upper right corner for each cell (grid in the y direction - rows).

  • nx (int)

    Number of nodes in the x direction (columns).

  • ny (int)

    Number of nodes in the y direction (rows).

  • point_a (float)

    First corner of the region of interest (ROI).

  • point_b (float)

    Corner opposed to point_b of the region of interest (ROI).

Methods:

define_mesh(point_a=None, point_b=None, nx=2, ny=2, show_grid=False)[source]

Method to construct the rectangular mesh used in the DIC analysis.

Input: * point_a (float)

First corner of the region of interest (ROI).

  • point_b (float)

    Corner opposed to point_b of the region of interest (ROI).

  • nx (int)

    Number of nodes in the x direction (columns), default 2.

  • ny (int)

    Number of nodes in the y direction (rows), default 2.

  • show_grid (int)

    This functionality is only available when point_a and point_b are provided. When noplot is True, the grid is not plotted for express calculations.

Output/Returns:

DICpy.DIC_2D.ImageRegistration

class DICpy.DIC_2D.ImageRegistration(mesh_obj=None)[source]

Bases: object

This class has the methods to perform the DIC analysis without considering the subpixel resolution. This is a parent class for children classes implementing methods with subpixel resolution.

Input: * mesh_obj (object)

Object of the RectangularMesh class.

Attributes:

  • pixel_dim (float)

    Size of each pixel in length dimension.

  • mesh_obj (object)

    Object of the RectangularMesh class.

  • u (ndarray)

    Displacements in the x (columns) dimension at the center of each cell.

  • v (ndarray)

    Displacements in the y (rows) dimension at the center of each cell.

Methods:

run()[source]

Method to perform the DIC analysis.

Input: * sub_pixel (str)

Method for the sub-pixel refining: - None: do not perform the sub-pixel refining. - ‘Crude’: crude method based on the image oversampling. - ‘gradient’: gradient based sub-pixel refining. - ‘coarse_fine’: method based on the sequential refining of the pixel domain. - ‘lukas_kanade’: method based on the Lukas-Kanade optical flow.

  • oversampling_x (int)

    Oversampling in the x dimension used when ‘Crude’ method is adopted, default is 1 (equal to ‘crude’).

  • oversampling_y (int)

    Oversampling in the y dimension used when ‘Crude’ method is adopted, default is 1 (equal to ‘crude’)..

  • niter (int)

    Number of iterations used when ‘coarse_fine’ method is adopted, default is 1.

Output/Returns:

DICpy.DIC_2D.Oversampling

class DICpy.DIC_2D.Oversampling(mesh_obj=None, over_x=1, over_y=1)[source]

Bases: DICpy.DIC_2D._image_registration.ImageRegistration

DIC with subpixel resolution using oversampling.

Input: * mesh_obj (object)

Object of the RegularGrid class.

Attributes:

  • pixel_dim (float)

    Size of each pixel in length dimension.

  • mesh_obj (object)

    Object of the RegularGrid class.

  • u (ndarray)

    Displacements in the x (columns) dimension at the center of each cell.

  • v (ndarray)

    Displacements in the y (rows) dimension at the center of each cell.

Methods:

DICpy.DIC_2D.GradientZero

class DICpy.DIC_2D.GradientZero(mesh_obj=None, max_iter=20, tol=0.001)[source]

Bases: DICpy.DIC_2D._image_registration.ImageRegistration

DIC with subpixel resolution using gradients. This class implement a method using zeroth order shape functions, and It is a child class of ImageRegistration.

Input: * mesh_obj (object)

Object of the RegularGrid class.

Attributes:

  • pixel_dim (float)

    Size of each pixel in length dimension.

  • mesh_obj (object)

    Object of the RegularGrid class.

  • u (ndarray)

    Displacements in the x (columns) dimension at the center of each cell.

  • v (ndarray)

    Displacements in the y (rows) dimension at the center of each cell.

Methods:

DICpy.DIC_2D.GradientOne

class DICpy.DIC_2D.GradientOne(mesh_obj=None, max_iter=20, tol=0.001)[source]

Bases: DICpy.DIC_2D._image_registration.ImageRegistration

DIC with subpixel resolution using gradients. This class implement a method using zeroth order shape functions, and It is a child class of ImageRegistration.

Input: * mesh_obj (object)

Object of the RegularGrid class.

Attributes:

  • pixel_dim (float)

    Size of each pixel in length dimension.

  • mesh_obj (object)

    Object of the RegularGrid class.

  • u (ndarray)

    Displacements in the x (columns) dimension at the center of each cell.

  • v (ndarray)

    Displacements in the y (rows) dimension at the center of each cell.

Methods:

DICpy.DIC_2D.CoarseFine

class DICpy.DIC_2D.CoarseFine(mesh_obj=None, niter=1)[source]

Bases: DICpy.DIC_2D._image_registration.ImageRegistration

DIC with subpixel resolution using an coarse-fine search approach, and It is a child class of ImageRegistration.

This method is presented in the following paper: “An advanced coarse-fine search approach for digital image correlation applications.” (By: Samo Simoncic, Melita Kompolsek, Primoz Podrzaj)

Input: * mesh_obj (object)

Object of the RegularGrid class.

Attributes:

  • pixel_dim (float)

    Size of each pixel in length dimension.

  • mesh_obj (object)

    Object of the RegularGrid class.

  • u (ndarray)

    Displacements in the x (columns) dimension at the center of each cell.

  • v (ndarray)

    Displacements in the y (rows) dimension at the center of each cell.

  • niter (int)

Number of iterations in the coarse-fine algorithm.

Methods:

DICpy.DIC_2D.LucasKanade

class DICpy.DIC_2D.LucasKanade(mesh_obj=None)[source]

Bases: DICpy.DIC_2D._image_registration.ImageRegistration

DIC with subpixel resolution using the Lucas-Kanade algorithm implemented in OpenCV-Python, and It is a child class of ImageRegistration.

Input: * mesh_obj (object)

Object of the RegularGrid class.

Attributes:

  • pixel_dim (float)

    Size of each pixel in length dimension.

  • mesh_obj (object)

    Object of the RegularGrid class.

  • u (ndarray)

    Displacements in the x (columns) dimension at the center of each cell.

  • v (ndarray)

    Displacements in the y (rows) dimension at the center of each cell.

Methods:

run()[source]

Method to run the DIC analysis.

DICpy.dic2d.PostProcessing

class DICpy.DIC_2D.PostProcessing(analysis_obj=None)[source]

Bases: object

This class contains the methods for visualizing the results of the DIC analysis. Input: * analysis_obj (object)

Object of the Analysis class.

Attributes:

  • analysis_obj (float)

    Object of the Analysis class.

  • mesh_obj (object)

    Object of the RectangularMesh class.

  • strain11 (ndarray)

    Strain xx at the center of each cell.

  • strain22 (ndarray)

    Strain yy at the center of each cell.

  • strain12 (ndarray)

    Strain xy at the center of each cell.

  • strain21 (ndarray)

    Strain yx (equal to Strain xy) at the center of each cell.

Methods:

get_fields()[source]

Method to estimate the strain fields.

Input:

Output/Returns:

visualization(results='u', step=0, smooth=False)[source]

Method to plot the results in terms of displacements and strains.

Input: * results (str)

Visulaize the results: -‘u’: displacement x. -‘v’: displacement y. -‘e11’: strain xx. -‘e22’: strain yy: -‘e12’: strain xy. -‘e21’: strain yx.

  • step (int)

    Define the result for a given loading step.

  • smooth (bool)

    Gaussian filtering.

Output/Returns:

DICpy.DIC_2D.Synthetic

class DICpy.DIC_2D.Synthetic(num_images=None, size=None, extension='png', random_state=None, save_images=True)[source]

Bases: object

This class contains the methods for generate synthetic data for the DIC analysis. Input:

Attributes:

  • path_speckle (str)

    Path to the speckle images.

  • path_calibration (str)

    Path to the calibration image.

  • ref_image (ndarray)

    Reference image.

  • calibration_image (ndarray)

    Calibration image.

  • images (ndarray)

    Gray images.

  • images_normalized (ndarray)

    Gray images normalized by 255.

  • lx (int)

    Number of columns of each image.

  • ly (int)

    Number of rows of each image.

  • num_img (int)

    Number of images.

  • pixel_dim (float)

    Size of each pixel in length dimension.

Methods:

generate_images(num_speckles=100, sigma=2, displacement_x=None, displacement_y=None)[source]