Image.hitormiss#

Image.hitormiss(s1: ndarray, s2: ndarray | None = None, border: str = 'replicate', bordervalue: float = 0, **kwargs: Any) → Self[source]#

Hit or miss transform

Parameters:

  • s1 (ndarray(N,M)) – structuring element 1

  • s2 (ndarray(N,M)) – structuring element 2

  • kwargs – arguments passed to opencv.morphologyEx

Returns:

transformed image

Return type:

Image

Return the hit-or-miss transform of the binary image which is defined by two structuring elements structuring elements

\[Y = (X \ominus S_1) \cap (X \ominus S_2)\]

which is the logical-and of the binary image and its complement, eroded by two different structuring elements. This preserves pixels where ones in the window are consistent with \(S_1\) and zeros in the window are consistent with \(S_2\).

If only s1 is provided it has three possible values:

  • 1, must match a non-zero value

  • -1, must match a zero value

  • 0, don’t care, matches any value.

Example:

>>> from machinevisiontoolbox import Image
>>> import numpy as np
>>> pixels = np.array([[0,0,1,0,1,1],[1,1,1,1,0,1],[0,1,0,1,1,0],[1,1,1,1,0,0],[0,1,1,0,1,0]])
>>> img = Image(pixels)
>>> img.print()
   0 0 1 0 1 1
   1 1 1 1 0 1
   0 1 0 1 1 0
   1 1 1 1 0 0
   0 1 1 0 1 0
>>> se = np.array([[0,1,0],[1,-1,1],[0,1,0]])
>>> se
array([[ 0,  1,  0],
       [ 1, -1,  1],
       [ 0,  1,  0]])
>>> img.hitormiss(se).print()
   0 0 0 1 0 0
   0 0 0 0 1 0
   1 0 1 0 0 0
   0 0 0 0 0 0
   1 0 0 1 0 0

Note

For the single argument case s1 \(=S_1 - S_2\).

References:

Important

Uses OpenCV function cv2.morphologyEx (with MORPH_HITMISS) which accepts single-channel, CV_8U or CV_16S images.

Seealso:

thin endpoint triplepoint opencv.morphologyEx