# -*- coding: utf-8 -*-
"""
Parameteric rank filters
This module provides rank filters that depend on several parameters.
"""
__author__ = 'Christoph Kirst <christoph.kirst.ck@gmail.com>'
__license__ = 'GPLv3 - GNU General Pulic License v3 (see LICENSE.txt)'
__copyright__ = 'Copyright © 2020 by Christoph Kirst'
__webpage__ = 'http://idisco.info'
__download__ = 'http://www.github.com/ChristophKirst/ClearMap2'
import numpy as np
from . import Rank as rnk
import pyximport;
pyximport.install(setup_args={"include_dirs":np.get_include()}, reload_support=True)
from . import ParametricCode as code
__all__ = ['nilblack', 'sauvola', 'contrast_limited_percentile', 'light_sheet_artifact_correction']
[docs]def nilblack(source, selem=None, sink=None, mask=None, k=0.0, **kwargs):
"""Local nilblack threshold of an image.
The resulting binary mask is True if the greyvalue of the center pixel is
greater than the local mean.
Arguments
---------
source : array
Input array.
selem : array
Structure element. If None, use a cube of size 3.
sink : array
Output array. If None, a new array is allocated.
mask : array
Optional mask, if None, the complete source is used.
Pixels on the mask are zero in the output.
max_bin : int or None
Maximal number of bins.
k : float
The factor between mean and standard deviation: threshold = mu + k * std
Returns
-------
sink : array
The filtered array.
"""
return rnk._apply_code(code.nilblack, code.nilblack_masked, sink_dtype = bool,
source=source, selem=selem, sink=sink, mask=mask, parameter_float=[k], **kwargs);
[docs]def sauvola(source, selem=None, sink=None, mask=None, k=0.0, R=1.0):
"""Local Sauvola threshold of an image.
The resulting binary mask is True if the greyvalue of the center pixel is
greater than the local mean.
Arguments
---------
source : array
Input array.
selem : array
Structure element. If None, use a cube of size 3.
sink : array
Output array. If None, a new array is allocated.
mask : array
Optional mask, if None, the complete source is used.
Pixels on the mask are zero in the output.
max_bin : int or None
Maximal number of bins.
k, R : float
The factors in the formula: threshold = mu + k * (std/R - 1) * (mu - max_bin)
Returns
-------
sink : array
The filtered array.
"""
return rnk._apply_code(code.sauvola, code.sauvola_masked, sink_dtype = None,
source=source, selem=selem, sink=sink, mask=mask, parameter_float=[k,R]);
[docs]def contrast_limited_percentile(source, selem=None, sink=None, mask=None, percentile=0.0, limit = None, contrast_limit = 0.1, **kwargs):
"""Local Sauvola threshold of an image.
The resulting binary mask is True if the greyvalue of the center pixel is
greater than the local mean.
Arguments
---------
source : array
Input array.
selem : array
Structure element. If None, use a cube of size 3.
sink : array
Output array. If None, a new array is allocated.
mask : array
Optional mask, if None, the complete source is used.
Pixels on the mask are zero in the output.
max_bin : int or None
Maximal number of bins.
k, R : float
The factors in the formula: threshold = mu + k * (std/R - 1) * (mu - max_bin)
Returns
-------
sink : array
The filtered array.
"""
if limit is None:
selem = rnk._initialize_selem(selem, source.ndim);
n_selem = np.sum(selem);
limit = int(np.round(contrast_limit * n_selem));
parameter_float = [percentile];
parameter_index = np.hstack([[limit], np.zeros(max_bin, dtype=int)]);
return rnk._apply_code(code.clp, code.clp_masked, sink_dtype = float,
source=source, selem=selem, sink=sink, mask=mask,
parameter_index=parameter_index, parameter_float=parameter_float, **kwargs);
[docs]def light_sheet_artifact_correction(source, selem=None, sink=None, mask=None, percentiles = (0.0, 1.0), limit = None, contrast_limit = 0.5, **kwargs):
"""Local Sauvola threshold of an image.
The resulting binary mask is True if the greyvalue of the center pixel is
greater than the local mean.
Arguments
---------
source : array
Input array.
selem : array
Structure element. If None, use a cube of size 3.
sink : array
Output array. If None, a new array is allocated.
mask : array
Optional mask, if None, the complete source is used.
Pixels on the mask are zero in the output.
max_bin : int or None
Maximal number of bins.
k, R : float
The factors in the formula: threshold = mu + k * (std/R - 1) * (mu - max_bin)
Returns
-------
sink : array
The filtered array.
"""
if limit is None:
selem = rnk._initialize_selem(selem, source.ndim);
n_selem = np.sum(selem);
limit = int(np.round(contrast_limit * n_selem));
parameter_float = np.array(percentiles, dtype = float);
parameter_index = np.hstack([[limit], np.zeros(max_bin, dtype=int)]);
return rnk._apply_code(code.lsac, code.lsac_masked, sink_dtype = float,
source=source, selem=selem, sink=sink, mask=mask,
parameter_index=parameter_index, parameter_float=parameter_float, **kwargs);
Source
