machinevisiontoolbox.Image.scalespace

Image.scalespace(n, sigma=1)

Compute image scalespace sequence

Parameters:

  • n (omt) – number of steps

  • sigma (scalar, optional) – Gaussian filter width, defaults to 1

Returns:

Gaussian and difference of Gaussian sequences, scale factors

Return type:

list of Image, list of Image, list of float

Compute a scalespace image sequence by consecutively smoothing the input image with a Gaussian of width sigma. The difference between consecutive smoothings is the difference of Gaussian which is an approximation to the Laplacian of Gaussian.

Examples:

>>> mona = Image.Read("monalisa.png", dtype="float");
>>> G, L, scales = mona.scalespace(8, sigma=8);
Note:

The two image sequences have the same length, the original image is not included in the list of smoothed images.

References:

  • Robotics, Vision & Control for Python, Section 12.3.2, P. Corke, Springer 2023.

Seealso:

pyramid smooth Kernel.Gauss Kernel.LoG