slmsuite.holography.analysis.blob_array_detect

blob_array_detect(img, size, orientation=None, orientation_check=True, dft_threshold=50, dft_padding=0, plot=False)[source]

Detect an array of spots and return the orientation as an affine transformation. Primarily used for calibration.

For a rectangular array of spots imaged in img, find the variables \(\vec{M}\) and \(\vec{b}\) for the affine transformation

\[\vec{y} = M \cdot \vec{x} + \vec{b}\]

which converts spot indices \(\vec{x}\) into camera pixel indices \(\vec{y}\).

Parameters

  • img (numpy.ndarray) – The image in question.

  • size ((int, int) OR int) – The size of the rectangular array in number of spots (Nx, Ny). If a single int size is given, then assume (N, N).

  • orientation (dict or None) – Guess array orientation (same format as the returned) from previous known results. If None (the default), orientation is estimated from looking for peaks in the Fourier transform of the image.

  • orientation_check (bool) – If enabled, looks for two missing spots at one corner as a parity check on rotation. Used by fourier_calibrate(). See make_rectangular_array().

  • dft_threshold (float in [0, 255]) – Minimum value of peak in blob detect of the DFT of img when orientation is None. Passed as kwarg to blob_detect() with name minThreshold.

  • dft_padding (int) – Increases the dimensions of the padded img before the DFT is taken when orientation is None. Dimensions are increased by a factor of 2 ** dft_padding. Increasing this value increases the k-space resolution of the DFT, and can improve orientation detection.

  • plot (bool) – Whether or not to plot debug plots. Default is False.

Returns

Orientation dictionary with the following keys, corresponding to the affine transformation:

  • "M" : numpy.ndarray (2, 2).

  • "b" : numpy.ndarray (2, 1).

Return type

dict