ESBO-ETC/esbo_etc/classes/sensor/PixelMask.py

121 lines
5.7 KiB
Python
Raw Normal View History

2020-05-13 14:32:48 +02:00
import astropy.units as u
from logging import warning
from ...lib.helpers import error, rasterizeCircle
import numpy as np
class PixelMask(np.ndarray):
"""
A class for modelling the pixel exposure mask for a pixel array.
"""
@u.quantity_input(pixel_geometry=u.pix, pixel_size="length", center_offset=u.pix)
def __new__(cls, pixel_geometry: u.Quantity, pixel_size: u.Quantity, center_offset: u.Quantity):
"""
Create a new pixel mask. Each coordinate is now converted to a index representation (y, x).
Parameters
----------
pixel_geometry : u.Quantity
The geometry of the pixel array in pixels [x, y]
pixel_size : length-Quantity
The edge length of a pixel (assumed to be square).
center_offset : u.Quantity
The offset of the PSF-center relative to the center of the detector array as length-quantity with two
entries: [offset in x-direction, offset in y-direction]
"""
# Create the ndarray instance of our type, given the usual
# ndarray input arguments. This will call the standard
# ndarray constructor, but return an object of our type.
# It also triggers a call to PixelMask.__array_finalize__
obj = super(PixelMask, cls).__new__(cls, (int(pixel_geometry.value[0]), int(pixel_geometry.value[1])),
dtype=float, buffer=None, offset=0, strides=None, order=None)
obj[:, :] = 0
# set the new attributes to the values passed
obj.pixel_geometry = [pixel_geometry[1], pixel_geometry[0]]
obj.pixel_size = pixel_size
obj.center_ind = [pixel_geometry[1].value / 2 - 0.5, pixel_geometry[0].value / 2 - 0.5]
obj.psf_center_ind = [obj.center_ind[0] + center_offset[1].value, obj.center_ind[1] + center_offset[0].value]
# Finally, we must return the newly created object:
return obj
def __array_finalize__(self, obj):
# ``self`` is a new object resulting from
# ndarray.__new__(PixelMask, ...), therefore it only has
# attributes that the ndarray.__new__ constructor gave it -
# i.e. those of a standard ndarray.
#
# We could have got to the ndarray.__new__ call in 3 ways:
# From an explicit constructor - e.g. PixelMask():
# obj is None
# (we're in the middle of the InfoArray.__new__
# constructor, and self.pixel_geometry will be set when we return to
# PixelMask.__new__)
if obj is None:
return
# From view casting - e.g arr.view(PixelMask):
# obj is arr
# (type(obj) can be PixelMask)
# From new-from-template - e.g mask[:3]
# type(obj) is PixelMask
#
# Note that it is here, rather than in the __new__ method,
# that we set the default value for our attributes, because this
# method sees all creation of default objects - with the
# PixelMask.__new__ constructor, but also with
# arr.view(PixelMask).
self.pixel_geometry = getattr(obj, 'pixel_geometry', None)
self.pixel_size = getattr(obj, 'pixel_size', None)
self.center_ind = getattr(obj, 'center_ind', None)
self.psf_center_ind = getattr(obj, 'psf_center_ind', None)
# We do not need to return anything
@u.quantity_input(radius=u.pix, center_offset=u.pix)
def createPhotometricAperture(self, shape: str, radius: u.Quantity, center_offset: u.Quantity = None):
"""
Create a photometric aperture on the pixel mask.
Parameters
----------
shape : str
Shape of the photometric aperture. This can be either 'circle' or 'square'.
radius : u.Quantity
The radius of the photometric aperture in pixels. In case of a square, the radius equals the half of the
side length.
center_offset : u.Quantity
The offset of the photometric aperture's centre with respect to the array's centre in pixels [x ,y]. The
origin of the coordinate system is in the upper left corner.
Returns
-------
"""
# Calculate the center coordinates
if center_offset:
xc = self.pixel_geometry[1] / 2 - 0.5 * u.pix + center_offset[0]
yc = self.pixel_geometry[0] / 2 - 0.5 * u.pix + center_offset[1]
else:
xc = self.psf_center_ind[1] * u.pix
yc = self.psf_center_ind[0] * u.pix
if (xc + radius).value > self.pixel_geometry[0].value - 1 or (xc - radius).value < 0 or\
(yc + radius).value > self.pixel_geometry[1].value - 1 or (yc - radius).value < 0:
warning("Some parts of the photometric aperture are outside of the array.")
if shape.lower() == "circle":
# Rasterize a circle on the grid
rasterizeCircle(self, radius.value, xc.value, yc.value)
elif shape.lower() == "square":
# Rasterize a square on the grid
# Calculate the left, right, upper and lower bounds of the square
x_right = int(round((xc + radius).value))
if x_right > self.pixel_geometry[0].value - 1:
x_right = self.pixel_geometry[0].value - 1
x_left = 0 if (xc - radius).value < 0 else int(round((xc - radius).value))
y_low = int(round((yc + radius).value))
if y_low > self.pixel_geometry[1].value - 1:
y_low = self.pixel_geometry[1].value - 1
y_up = 0 if (yc - radius).value < 0 else int(round((yc - radius).value))
# Mark the pixels contained in the square with 1
self[y_up:(y_low + 1), x_left:(x_right + 1)] = 1
else:
error("Unknown photometric aperture shape: '" + shape + "'.")