researchutils
master

Contents:

  • API Reference
    • Utilities related to arrays
    • Utilities related to chainer
    • Utilities related to files
    • Utilities related to images
      • Converter
      • Preprocess
      • Viewer
    • Utilities related to math
researchutils
  • Docs »
  • API Reference »
  • Utilities related to images
  • Edit on GitHub

Utilities related to images¶

Converter¶

researchutils.image.converter.chw2hwc(image)[source]¶

Changes the order of image pixels from Color-Height-Width to Height-Width-Color

Parameters:image (numpy.ndarray) – Image with pixels in Color-Height-Width order
Returns:image – Image with pixels in Height-Width-Color order
Return type:numpy.ndarray
researchutils.image.converter.hwc2chw(image)[source]¶

Changes the order of image pixels from Height-Width-Color to Color-Height-Width

Parameters:image (numpy.ndarray) – Image with pixels in Height-Width-Color order
Returns:image – Image with pixels in Color-Height-Width order
Return type:numpy.ndarray

Preprocess¶

researchutils.image.preprocess.compute_mean_image(frames)[source]¶

Compute mean image of given frames

Parameters:frames (list of numpy.ndarray) – All frame must have same shape
Returns:mean_image – Mean image of given frame.
Return type:numpy.ndarray

Viewer¶

researchutils.image.viewer.animate(images, comparisons=None, titles=[], is_gray=False, fps=15, repeat=False, save_gif=False, save_mp4=False, auto_close=False)[source]¶

Animate list of images

Parameters:
  • images (list of numpy.ndarray) – Images to display
  • comparisons (list of numpy.ndarray) – Comparison images or multiple list of comparison images to display (Optional)
  • titles (list of string) – Titles for each animation
  • is_gray (bool) – Whether given images are grayscale or colored True if images are grayscale
  • fps (int) – Number of frames to display in 1 second. Abbreviation of “frame per second”
  • repeat (bool) – Repeat animation when reaches end of images
  • save_gif (bool) – Save animation as gif image
  • save_mp4 (bool) – Save animation as mp4
  • auto_close (bool) – Close animation window after finish animating
researchutils.image.viewer.animate_in_matrix_form(images=None, titles=[], is_gray=False, fps=15, images_per_row=None, repeat=False, save_gif=False, save_mp4=False, auto_close=False)[source]¶

Animate list of images in matrix form

Parameters:
  • images (list of numpy.ndarray) – Images to display
  • titles (list of string) – Titles for each animation
  • is_gray (bool) – Whether given images are grayscale or colored True if images are grayscale
  • fps (int) – Number of frames to display in 1 second. Abbreviation of “frame per second”
  • images_per_row (int) – Number of images to display in each row
  • repeat (bool) – Repeat animation when reaches end of images
  • save_gif (bool) – Save animation as gif image
  • save_mp4 (bool) – Save animation as mp4
  • auto_close (bool) – Close animation window after finish animating
researchutils.image.viewer.show_image(image, title='', is_gray=False)[source]¶

Display given image

Parameters:
  • image (numpy.ndarray) – Image to display
  • title (string) – Title of the image to display
  • is_gray (bool) – Whether given image is grayscale or colored True if image is grayscale
researchutils.image.viewer.show_images(images, title='', comparisons=None, comparison_title='', is_gray=False, vertical=False)[source]¶

Display given image

Parameters:
  • images (list of numpy.ndarray) – Images to display
  • title (string) – Title of each image to display. If length of images is greater than 1, index of each image will be appended to the tail of title.
  • comaprisons (list of numpy.ndarray) – Comparison images to display
  • comaprison_title (list of numpy.ndarray) – Title of each comparison image to display. If length of comparisons is greater than 1, index of each image will be appended to the tail of title.
  • is_gray (bool) – Whether given images are grayscale or colored True if images are grayscale
  • vertical (bool) – Display comparisons vertically
Next Previous

© Copyright 2018, Yu Ishihara Revision 34a5e71f.

Built with Sphinx using a theme provided by Read the Docs.