Image.pyramid#

Image.pyramid(sigma: int | float = 1, N: int | None = None, border: str = 'replicate', bordervalue: int | float = 0) → list[Any][source]#

Pyramidal image decomposition

Parameters:

  • sigma (float) – standard deviation of Gaussian kernel

  • N (int, optional) – number of pyramid levels to be computed, defaults to all

  • border (str, optional) – option for boundary handling, see convolve, defaults to ‘replicate’

  • bordervalue (scalar, optional) – padding value, defaults to 0

Returns:

list of images at each pyramid level

Return type:

list of Image

Returns a pyramid decomposition of the input image using Gaussian smoothing with standard deviation of sigma. The return is a list array of images each one having dimensions half that of the previous image. The pyramid is computed down to a non-halvable image size.

Example:

>>> from machinevisiontoolbox import Image
>>> img = Image.Read('monalisa.png')
>>> pyramid = img.pyramid(4)
>>> len(pyramid)
11
>>> pyramid
[Image(size=(677, 700), dtype=uint8), Image(size=(339, 350), dtype=uint8), Image(size=(170, 175), dtype=uint8), Image(size=(85, 88), dtype=uint8), Image(size=(43, 44), dtype=uint8), Image(size=(22, 22), dtype=uint8), Image(size=(11, 11), dtype=uint8), Image(size=(6, 6), dtype=uint8), Image(size=(3, 3), dtype=uint8), Image(size=(2, 2), dtype=uint8), Image(size=(1, 1), dtype=uint8)]

Note

  • Works for greyscale images only.

  • Converts a color image to greyscale.

References:

Important

Uses OpenCV function cv2.pyrDown which accepts single-channel, CV_8U, CV_16U, CV_16S, CV_32F or CV_64F images (color images are automatically converted to greyscale).

Seealso:

smooth scalespace cv2.pyrDown