Module functions

Functions have been documented using sphinx.ext.autodoc and the docstring style extension ‘sphinx.ext.napoleon’:

Module contours

Modules.contours.al_ratio(contour_length, contour_area)

Determines ratio of contour length and area

Parameters

  • contour_length (ndarray) – 1 dimensional array with type float that contains length of contours

  • contour_area (ndarray) – 1 dimensional array with type float that contains area of contours

Returns

ratio – 1 dimensional array with type float that contains ratio of contour length to area

Return type

ndarray

Modules.contours.draw_cont(img, contour, name, directory)

Draws contours and returns image

Parameters

  • img (ndarray) – N dimensional array with type int that contains the image to process

  • contour (contours) – Contours that contains list of coordinates of detected contours

  • name (str) – The name of the returned image

  • directory (str) – The directory where the image is saved

Returns

new_img – N x N dimensional array that contains the selected contours

Return type

ndarray

Modules.contours.filter_cont(contour, index)

Selects contours by index

Parameters

  • contour (contours) – Contours that contains list of coordinates of detected contours

  • index (ndarray) – 1 dimensional array with type int that contains index of selected contours

Returns

cnt – N x 2 dimensional array with type contours that contains selected contours

Return type

ndarray

Modules.contours.filter_df(ind, c_length, c_area, c_ratio)

Writes new DataFrame with selected data

Parameters

  • ind (ndarray) – 1 dimensional array with type int that contains the index of data selected from other DataFrame

  • c_length (ndarray) – 1 dimensional array with type float that contains the length of selected contours

  • c_area (ndarray) – 1 dimensional array with type float that contains the area of selected contour

  • c_ratio (ndarray) – 1 dimensional array with type float that contains the ratio of contour length and area

Returns

  • df (DataFrame) – DataFrame that contains properties of selected contours

  • length (ndarray) – 1 dimensional array with type float that contains the length of each selected contour

  • area (ndarray) – 1 dimensional array with type float that contains the area of each selected contour

  • ratio (ndarray) – 1 dimensional array with type float that contains the length/area ratio of each selected contour

Modules.contours.filter_values(df, column, min_val, max_val)

Filters column in DataFrame by values

Parameters

  • df (DataFrame) – DataFrame that contains contour properties

  • column (string) – The name of the column

  • min_val (float) – The minimum value of the data

  • max_val (float) – The maximum value of the data

Returns

ind – 1 dimensional array with type int that describes values, which fulfill conditions

Return type

ndarray

Modules.contours.find_cont(img)

Detects contours and draws them in RGB image

Parameters

img (ndarray) – N x N dimensional array with type int that contains the image to process

Returns

  • canvas (ndarray) – N x N dimensional array that contains the detected contours

  • contour (contours) – List of coordinates of contours that were detected in the image

Modules.contours.find_contour_img(img, img_thr, directory)

Finds contours, filters them by manually chosen values and returns filtered contours

Parameters

  • img (ndarray) – N x N dimensional array with type int that contains the original image

  • img_thr (ndarray) – N x N dimensional array with type int that contains the thresholded image

  • directory (str) – The directory where the image is saved

Returns

c_a – N x N x 2 array with type contours that contains selected contours

Return type

ndarray

Modules.contours.find_extrema(img)

Finds index of max and min value in image

Parameters

img (image) – N dimensional array with type float that contains the z_values of the data

Returns

  • z_min (int) – Minimum value in array

  • z_max (int) – Maximum value in array

Modules.contours.get_contarea(contour)

Determines area that is surrounded by contour and its length

Parameters

contour (contours) – Contours that contains list of coordinates of detected contours

Returns

  • length (ndarray) – 1 dimensional array with type float that contains length of contours

  • area (ndarray) – 1 dimensional array with type float that contains area of contours

Modules.contours.normalise_zvalues(img)

Returns normalised grayscale image

Parameters

img (image) – N dimensional array with type float that contains the z_values of the data

Returns

imggr – N x N dimensional array with type int that contains grayscale values of the image

Return type

ndarray

Modules.contours.RGB_2_bin(img, value)

Converts RGB image to grayscale image

Parameters

  • img (image) –

  • value (int) – The value, which determines if the point is True or False

Returns

img_new – N x N dimensional array that contains a binary image

Return type

ndarray

Modules.contours.write_df(c_length, c_area, ratio)

Writes data in pandas data frame

Parameters

  • c_length (ndarray) – 1 dimensional array with type float that contains list of contour lengths

  • c_area (ndarray) – 1 dimensional array with type float that contains list of contour areas

  • ratio (ndarray) – 1 dimensional array with type float that contains ratio of contour length to area

Returns

df – DataFrame that contains contour length, area and ratio length/area

Return type

DataFrame

Module extractCuts

Modules.extractCuts.calc_asymcut(values, coord, ori, imgh, imgrh, imgw, imgrw, centre)

Determines the cross-section along the most asymmetric cut through the z_values and shifts the origin to the geometric centre

Parameters

  • values (ndarray) – N dimensional array with type float that contains the grayscale values of the original image

  • coord (ndarray) – 1 dimensional array with type (float, float), (float, float) that contains the x- and y-coordinate left of a point and the x- and y-coordinate right of a point

  • ori (str) – Orientation of the straight: parallel, perpendicular

  • imgh (int) – The height of the image

  • imgrh (int) – The resolution of the image in y-direction

  • imgw (int) – The width of the image

  • imgrw (int) – The resolution of the imag in x-direction

  • centre (ndarray) – N dimensional array with type float, float that contains the x- and y-coordinates of the geometric centre

Returns

  • cxpara (ndarray) – 1 dimensional array with type float that contains the x-coordinates of the asymmetric cut

  • cypara (ndarray) – 1 dimensional array with type float that contains the y-coordinates of the asymmetric cut

  • zpara (ndarray) – 1 dimensional array with type float that contains the z-coordinates of the asymmetric cut

  • cxperp (ndarray) – 1 dimensional array with type float that contains the x-coordinates of the perpendicular asymmetric cut

  • cyperp (ndarray) – 1 dimensional array with type float that contains the y-coordinates of the perpendicular asymmetric cut

  • ziperp (ndarray) – 1 dimensional array with type float that contains the z-coordinates of the perpendicular asymmetric cut

  • zipar (ndarray) – 1 dimensional array with type float that contains the z-coordinates of the asymmetric cut

  • cut (ndarray) – 1 dimensional array with type float that contains the points of the asymmetric cut in µm

  • pcut (ndarray) – 1 dimensional array with type float that contains the points of the perpendicular asymmetric cut in µm

  • x2cs (ndarray) – 2 dimensional array with type float that contains the x-coordinates of the asymmetric parallel and perpendicular cut

  • y2cs (ndarray) – 2 dimensional array with type float that contains the y-coordinates of the asymmetric parallel and perpendicular cut

  • z2cs (ndarray) – 2 dimensional array with type float that contains the z-coordinates of the asymmetric parallel and perpendicular cut

  • cut2 (ndarray) – 2 dimensional array with type float that contains the points of the parallel and perpendicular asymmetric cut in µm

Modules.extractCuts.calc_straight(circle_found, mcentre, centre, imgh, imgrh, imgw, imrw)

Determines the straight betweeen the centre of mass and the geometric centre and the values of the points on the straight within a range of 50 pixels (much larger than cell diameter) around the geometric centre. Calculates the perpendicular straight and the values of the points on the straight.

Parameters

  • circle_found (ndarray) – 1 dimensional array with type float, float, float that contains the x- and y-coordinate of the centre and the radius of the bounding circle

  • mcentre (ndarray) – N dimensional array with type float, float that contains the x- and y-coordinates of the centre of mass weighted by the corresponding z-values

  • centre (ndarray) – N dimensional array with type float, float that contains the x- and y-coordinates of the geometric centre

  • imgh (int) – The height of the image

  • imgrh (int) – The resolution of the image in y-direction

  • imgw (int) – The width of the image

  • imgrw (int) – The resolution of the imag in x-direction

Returns

  • xrs (ndarray) – 1 dimensional array with type float that contains all x coordinates on the straight right of the geometric centre within a range of 50 pixels

  • yrs (ndarray) – 1 dimensional array with type float that contains all y coordinates on the straight right of the geometric centre within a range of 50 pixels

  • xls (ndarray) – 1 dimensional array with type float that contains all x coordinates on the straight left of the geometric centre within a range of 50 pixels

  • yls (ndarray) – 1 dimensional array with type float that contains all y coordinates on the straight left of the geometric centre within a range of 50 pixels

  • xprs (ndarray) – 1 dimensional array with type float that contains all x coordinates on the perpendicular straight right of the geometric centre within a range of 50 pixels

  • yprs (ndarray) – 1 dimensional array with type float that contains all y coordinates on the perpendicular straight right of the geometric centre within a range of 50 pixels

  • xpls (ndarray) – 1 dimensional array with type float that contains all x coordinates on the perpendicular straight left of the geometric centre within a range of 50 pixels

  • ypls (ndarray) – 1 dimensional array with type float that contains all y coordinates on the perpendicular straight left of the geometric centre within a range of 50 pixels

Modules.extractCuts.centre_of_mass(all_coord, img)

Determines the centre of mass inside a rectangle weighted by the corresponding z-values

Parameters

  • all_coord (ndarray) – 1 dimensional array with type float, float that contains all x- and y-coordinates inside a rectangle

  • img (ndarray) – N dimensional array with type float that contains the z_values of the data

Returns

centre_mass – N x 2 array with type float, float that contains the x- and y-coordinates of the centre of mass weighted by the corresponding z-values

Return type

ndarray

Modules.extractCuts.find_coordborder(cnt_drawn, img)

Finds points inside cell contour and returns their coordinates

Parameters

  • cnt_drawn (ndarray) – 1 dimensional array with type int, int, float, float that contains the index of a rectangle with corresponding contour, the y- and x-coordinate of the contour

  • img (ndarray) – N dimensional array with type float that contains the z_values of the data

Returns

img_coord – N x 3 array with type float, float, float that contains the y-coordinate, the minimum and maximum x-coordinate of the points inside a contour

Return type

ndarray

Modules.extractCuts.geometric_centre(img_coord)

Determines the geometric centre of a contour (cell)

Parameters

img_coord (ndarray) – 1 dimensional array with type float, float, float that contains the y-coordinate, the minimum and maximum x-coordinate

Returns

  • centre (ndarray) – N x 2 array with type float, float that contains the x- and y-coordinate of the centre of a rectangle

  • all_coord (ndarray) – N x N x 2 array with type float, float that contains all x- and y-coordinates inside a rectangle

Modules.extractCuts.match_coord(x_straight, y_straight, cntCell)

Finds contour points on the straight and sorts the coordinates by distance of the point from the geometric centre

Parameters

  • x_straight (ndarray) – 1 dimensional array with type float that contains all x coordinates on the straight within a range of 50 pixels

  • y_straight (ndarray) – 1 dimensional array with type float that contains all x coordinates on the straight within a range of 50 pixels

  • cnt_drawn (ndarray) – 1 dimensional array with type int, int, float, float that contains the index of a rectangle with corresponding contour, the y- and x-coordinate of the contour

Returns

  • match_coord (ndarray) – N x 3 array with type float, float, float that contains the x-coordinate, the y-coordinate and the distance of the point from the geometric centre

  • sort_coord (ndarray) – N x 2 array with type float, float that contains the x- and y-coordinate of a point

Modules.extractCuts.points_incell(img_coord, img)

Determines the z-value of all points inside a contour

Parameters

  • img_coord (ndarray) – 1 dimensional array with type float, float, float that contains the y-coordinate, the minimum and maximum x-coordinate of the points inside a contour

  • img (ndarray) – N dimensional array with type float that contains the z_values of the data

Returns

points_inside – N x N dimensional array with type float that contains all z-values inside a contour

Return type

ndarray

Modules.extractCuts.sort_coord(coord_rmatch, coord_lmatch)

Sorts the matching coordinates by distance between left point and right point

Parameters

  • coord_rmatch (ndarray) – 2 dimensional array with type float, float that contains the x-coordinate and the y-coordinate right of the centre

  • coord_lmatch (ndarray) – 2 dimensional array with type float, float that contains the x-coordinate and the y-coordinate left of the centre

Returns

coord_sort – N x 2 dimensional array with type float, float that contains the x- and y-coordinates of the points intersecting with the straight of highest asymmetry left and right of the centre of the object

Return type

ndarray

Module NN

Modules.NN.format_values(name, spath)

Formats values and writes them into array

Parameters

  • name (str) – Name of the txt file

  • spath (str) – Name of the directory

  • parameter (str) – Name of the parameter

Returns

values – 1 dimensional array with type float that contains the content of the text file

Return type

ndarray

Modules.NN.format_NN_values(name, spath)

Formats values for NN input and writes them into array

Parameters

  • name (str) – Name of the txt file

  • spath (str) – Name of the directory

  • parameter (str) – Name of the parameter

Returns

values – 1 dimensional array with type float that contains the content of the text file

Return type

ndarray

Modules.NN.read_images(path)

Reads text files of characteristic cuts and labels them according to their category

Parameters

path (str) – The name of the folder that contains the characteristic cuts

Returns

  • carr_form (ndarray) – N x (N x N) dimensional array with type float that contains N times two characteristic cross-sections with 50 data points

  • label_form (ndarray) – 1 dimensional array with type int that contains the labels of N cells

Modules.NN.train_model(model, carr_form, label_form, cut_train, label_train, directory, mname, tlabel)

Trains NN model on a combination of old and new data

Parameters

  • model (Sequential) – The pre-trained neural network

  • carr_form (ndarray) – N x 2 x 50 array with type float that contains N times two characteristic cross-sections with 50 data points

  • label_form (ndarray) – N x 1 array with type int that contains the labels of N cells

  • directory (str) – The name of the network folder

  • mname (str) – The name of the new network

  • tlabel (tkinter Label) – Displays the progres of the calculation

Returns

model – Sequential class that contains the trained new keras model

Return type

tf.keras.Sequential