#!/usr/bin/env python
from collections import namedtuple
import matplotlib.pyplot as plt
import numpy as np
import scipy as sp
from scipy import interpolate
import cv2 as cv
from pathlib import Path
import os.path
from spatialmath.base import argcheck, getvector
from machinevisiontoolbox.base import (
iread,
iwrite,
colorname,
int_image,
float_image,
idisp,
)
import os
import cv2 as cv
import zipfile
import numpy as np
import fnmatch
from numpy.core.numeric import _rollaxis_dispatcher
from machinevisiontoolbox.base import mvtb_path_to_datafile, iread, convert
from numpy.lib.arraysetops import isin
[docs]class ImageIOMixin:
# ======================= image i/io ================================== #
[docs] @classmethod
def Read(cls, filename, alpha=False, rgb=True, **kwargs):
"""
Read image from file
:param filename: image file name
:type filename: str
:param alpha: include alpha plane if present, defaults to False
:type alpha: bool, optional
:param rgb: force color image to be in RGB order, defaults to True
:type rgb: bool, optional
:param kwargs: options applied to image frames, see :func:`~machinevisiontoolbox.base.imageio.convert`
:raises ValueError: file not found
:return: image from file
:rtype: :class:`Image`
Load monochrome or color image from file, many common formats are
supported. A number of transformations can be applied to the image
loaded from the file before it is returned.
Example:
.. runblock:: pycon
>>> from machinevisiontoolbox import Image
>>> Image.Read('street.png')
>>> Image.Read('flowers1.png')
>>> Image.Read('flowers1.png', grey=True)
>>> Image.Read('flowers1.png', dtype='float16')
>>> Image.Read('flowers1.png', reduce=4)
>>> Image.Read('flowers1.png', gamma='sRGB') # linear tristimulus values
:note: If the path is not absolute it is first searched for relative
to the current directory, and if not found, it is searched for in
the ``images`` folder of the ``mvtb_data`` package.
:seealso: :func:`~machinevisiontoolbox.base.imageio.iread` :func:`~machinevisiontoolbox.base.imageio.convert` `cv2.imread <https://docs.opencv.org/master/d4/da8/group__imgcodecs.html#ga288b8b3da0892bd651fce07b3bbd3a56>`_
"""
if not isinstance(filename, (str, Path)):
raise ValueError("expecting a string or path")
# read the image
data = iread(filename, rgb=rgb, **kwargs)
# result is a tuple(image, filename) or a list of tuples
colororder = None
if isinstance(data, tuple):
# singleton image, make it a list
image, name = data
if not alpha and image.ndim == 3 and image.shape[2] == 4:
image = image[:, :, :3]
if image.ndim > 2:
colororder = "RGB" if rgb else "BGR"
return cls(
image, name=name, colororder=colororder
) # OpenCV file read order)
elif isinstance(data, list):
raise ValueError("wildcard read not support, use FileCollection")
[docs] def disp(self, title=None, **kwargs):
"""
Display image
:param title: named of window, defaults to image ``name``
:type title: bool
:param kwargs: options, see :func:`~machinevisiontoolbox.base.imageio.idisp`
Display an image using either Matplotlib (default) or OpenCV.
Example:
.. runblock:: pycon
>>> from machinevisiontoolbox import Image
>>> img = Image.Read('flowers1.png')
>>> img.disp()
.. plot::
from machinevisiontoolbox import Image
Image.Read('flowers1.png').disp()
:seealso: :func:`~machinevisiontoolbox.base.imageio.idisp`
"""
if title is False:
title = None
elif title is None and self.name is not None:
_, title = os.path.split(self.name)
if self.domain is not None:
# left right top bottom
kwargs["extent"] = [
self.domain[0][0],
self.domain[0][-1],
self.domain[1][-1],
self.domain[1][0],
]
return idisp(
self.A, title=title, colororder="RGB" if self.isrgb else "BGR", **kwargs
)
[docs] def write(self, filename, dtype="uint8", **kwargs):
"""
Write image to file
:param filename: filename to write to
:type filename: str
:param dtype: data type to convert to, before writing
:type dtype: str
:param kwargs: options for :func:`~machinevisiontoolbox.base.iwrite`
Write image data to a file. The file format is taken from the extension
of the filename.
Example:
.. runblock:: pycon
>>> from machinevisiontoolbox import Image
>>> img = Image.Read('flowers1.png')
>>> img.write('flowers.jpg')
:seealso: :func:`~machinevisiontoolbox.base.iwrite` `cv2.imwrite <https://docs.opencv.org/master/d4/da8/group__imgcodecs.html#gabbc7ef1aa2edfaa87772f1202d67e0ce>`_
"""
# cv.imwrite can only save 8-bit single channel or 3-channel BGR images
# with several specific exceptions
# https://docs.opencv.org/4.4.0/d4/da8/group__imgcodecs.html
# #gabbc7ef1aa2edfaa87772f1202d67e0ce
# TODO imwrite has many quality/type flags
ret = iwrite(self.image.astype(dtype), filename, **kwargs)
return ret
[docs] def showpixels(
self, textcolors=["yellow", "blue"], fmt=None, ax=None, windowsize=0, **kwargs
):
"""
Display image with pixel values
:param textcolors: text color, defaults to ['yellow', 'blue']
:type textcolors: list, optional
:param fmt: format string for displaying pixel values, defaults to None
:type fmt: str, optional
:param ax: Matplotlb axes to draw into, defaults to None
:type ax: axes, optional
:param windowsize: half side length of superimposed moving window, defaults to 0
:type windowsize: int, optional
:return: a moving window
:rtype: ``Window`` instance
Display a monochrome image with the pixel values overlaid. This is
suitable for small images, of order 10x10, used for pedagogical
purposes. For example it can be used to animate the operation of
sliding window operations like convolution or morphology.
The first color in ``textcolors`` is used for pixels below 50% intensity
and the second color for those above 50%.
If ``windowsize`` is given then a translucent colored window is
superimposed and a ``Window`` instance returned. This allows the window
position, color and opacity to be changed.
Example:
.. runblock:: pycon
>>> from machinevisiontoolbox import Image
>>> img = Image.Random(10)
>>> window = img.showpixels(windowsize=1) # with 3x3 window
>>> window.move(2,3) # position window at (2,3)
>>> window.move(4,5, color='blue', alpha=0.7)
.. plot::
from machinevisiontoolbox import Image
img = Image.Random(10)
img.showpixels(windowsize=1)
.. plot::
from machinevisiontoolbox import Image
img = Image.Random(10)
window = img.showpixels(windowsize=1)
window.move(2,3)
:seealso: :meth:`print`
"""
if ax is None:
ax = plt.gca()
if self.isint:
fmt = "{:d}"
halfway = self.maxval / 2
else:
fmt = "{:.2f}"
halfway = 0.5
image = self.image
for v in range(self.height):
for u in range(self.width):
if isinstance(textcolors, (list, tuple)):
if image[v, u] < halfway:
color = textcolors[0]
else:
color = textcolors[1]
elif textcolors == "grey":
if image[v, u] < halfway:
color = image[v, u] + 0.4 * np.r_[1, 1, 1]
else:
color = image[v, u] - 0.4 * np.r_[1, 1, 1]
ax.text(
u,
v,
fmt.format(image[v, u]),
horizontalalignment="center",
verticalalignment="center",
color=color,
**kwargs,
)
ax.imshow(image, cmap="gray")
ax.set_xlabel("u (pixels)")
ax.set_ylabel("v (pixels)")
plt.draw()
class Window:
def __init__(self, h=1, color="red", alpha=0.6, ax=None):
self.h = h
self.color = color
self.alpha = alpha
w = 2 * h + 1
patch = plt.Rectangle((0, 0), w, w, color=color, alpha=alpha)
if ax is None:
ax = plt.gca()
ax.add_patch(patch)
self.patch = patch
def move(self, u, v, color=None, alpha=0.5):
if color is not None:
self.color = color
self.patch.set_color(color)
if alpha is not None:
self.alpha = alpha
self.patch.set_alpha(alpha)
self.patch.set_x(u - self.h - 0.5)
self.patch.set_y(v - self.h - 0.5)
if windowsize > 0:
return Window(windowsize)
# def ascvtype(self):
# if np.issubdtype(self.image.dtype, np.floating):
# return self.image.astype(np.float32)
# else:
# return self.image.astype(np.uint8)
[docs] def anaglyph(self, right, colors="rc", disp=0):
"""
Convert stereo images to an anaglyph image
:param right: right image
:type right: Image instance
:param colors: lens colors (left, right), defaults to 'rc'
:type colors: str, optional
:param disp: disparity, defaults to 0
:type disp: int, optional
:raises ValueErrror: images are not the same size
:return: anaglyph image
:rtype: :class:`Image`
Returns an anaglyph image which combines the two images of a stereo pair
by coding them in two different colors. By default the left image is
red, and the right image is cyan.
``colors`` describes the lens color coding as a string with 2 letters,
the first for left, the second for right, and each is one of:
==== ========
code color
==== ========
'r' red
'g' green
'b' green
'c' cyan
'm' magenta
==== ========
If ``disp`` is positive the disparity is increased by shifting the
``right`` image to the right. If negative disparity is reduced by
shifting the ``right`` image to the left. These adjustments are
achieved by trimming the images. Use this option to make the images
more natural/comfortable to view, useful if the images were captured
with a stereo baseline significantly different to the human eye
separation (typically 65mm).
Example:
.. runblock:: pycon
>>> from machinevisiontoolbox import Image
>>> left = Image.Read("rocks2-l.png", reduce=2)
>>> right = Image.Read("rocks2-r.png", reduce=2)
>>> left.anaglyph(right).disp()
.. plot::
from machinevisiontoolbox import Image
left = Image.Read("rocks2-l.png", reduce=2)
right = Image.Read("rocks2-r.png", reduce=2)
left.anaglyph(right).disp()
:reference:
- Robotics, Vision & Control for Python, Section 14.4, P. Corke, Springer 2023.
:seealso: :meth:`stdisp` :meth:`Overlay`
"""
if self.size != right.size:
raise ValueError("images must be same size")
width, height = self.size
# ensure the images are greyscale
left = self.mono()
right = right.mono()
if disp > 0:
left = left.trim(right=disp)
right = right.trim(left=disp)
elif disp < 0:
disp = -disp
left = left.trim(left=disp)
right = right.trim(right=disp)
colordict = {
"r": (1, 0, 0),
"g": (0, 1, 0),
"b": (0, 0, 1),
"c": (0, 1, 1),
"m": (1, 0, 1),
"o": (1, 1, 0),
}
return left.colorize(colordict[colors[0]]) + right.colorize(
colordict[colors[1]]
)
[docs] def stdisp(self, right):
"""
Interactive display of stereo image pair
:param right: right image
:type right: :class:`Image`
The left and right images are displayed, stacked horizontally. Clicking
in the left-hand image sets a crosshair cursor in the right-hand
image. Clicking the corresponding point in the right-hand image
will display the disparity at the top of the right-hand image.
Example::
>>> from machinevisiontoolbox import Image
>>> left = Image.Read("rocks2-l.png", reduce=2)
>>> right = Image.Read("rocks2-r.png", reduce=2)
>>> left.stdisp(right)
.. plot::
from machinevisiontoolbox import Image
left = Image.Read("rocks2-l.png", reduce=2)
right = Image.Read("rocks2-r.png", reduce=2)
left.stdisp(right)
:note: The images are assumed to be epipolar aligned.
:reference:
- Robotics, Vision & Control for Python, Section 14.4, P. Corke, Springer 2023.
:seealso: :meth:`anaglyph`
"""
class Cursor:
"""
A cross hair cursor.
"""
def __init__(self, ax, ax2):
self.ax = ax
self.ax2 = ax2
self.horizontal_line = ax.axhline(color="k", lw=0.8)
self.horizontal_line2 = ax2.axhline(color="k", lw=0.8)
self.vertical_line = ax.axvline(color="k", lw=0.8)
self.vertical_line2 = ax2.axvline(color="k", lw=0.8)
self.vertical_line3 = ax2.axvline(color="k", lw=0.8, ls="--")
self.leftclicked = False
self.x_left = None
# text location in axes coordinates
self.text = self.ax2.text(
0.05, 0.95, "", transform=ax2.transAxes, backgroundcolor="w"
)
def set_cross_hair_visible(self, visible):
need_redraw = self.horizontal_line.get_visible() != visible
self.horizontal_line.set_visible(visible)
self.horizontal_line2.set_visible(visible)
self.vertical_line.set_visible(visible)
self.vertical_line2.set_visible(visible)
self.vertical_line3.set_visible(visible)
self.text.set_visible(visible)
return need_redraw
def on_mouse_move(self, event):
if event.inaxes == self.ax2 and self.leftclicked:
x, y = event.xdata, event.ydata
# update the line positions
self.vertical_line3.set_xdata(x)
self.text.set_text("d={:.2f}".format(self.x_left - x))
self.ax2.figure.canvas.draw()
# if event.inaxes:
# need_redraw = self.set_cross_hair_visible(False)
# if need_redraw:
# self.ax.figure.canvas.draw()
# else:
# self.set_cross_hair_visible(True)
# x, y = event.xdata, event.ydata
# # update the line positions
# self.horizontal_line.set_ydata(y)
# self.vertical_line.set_xdata(x)
# # self.text.set_text('x=%1.2f, y=%1.2f' % (x, y))
# self.ax.figure.canvas.draw()
def on_click(self, event):
# if not event.inaxes:
# need_redraw = self.set_cross_hair_visible(False)
# if need_redraw:
# self.ax.figure.canvas.draw()
# else:
if event.inaxes == self.ax:
self.set_cross_hair_visible(True)
x, y = event.xdata, event.ydata
# update the line positions
self.horizontal_line.set_ydata(y)
self.vertical_line.set_xdata(x)
self.horizontal_line2.set_ydata(y)
self.vertical_line2.set_xdata(x)
self.ax.figure.canvas.draw()
self.leftclicked = True
self.x_left = x
fig, (ax1, ax2) = plt.subplots(ncols=2, sharey=True)
self.disp(ax=ax1, grid=True)
right.disp(ax=ax2, grid=True)
cursor = Cursor(ax1, ax2)
fig.canvas.mpl_connect("motion_notify_event", cursor.on_mouse_move)
fig.canvas.mpl_connect("button_press_event", cursor.on_click)
plt.show(block=True)
# --------------------------------------------------------------------------- #
if __name__ == "__main__":
import pathlib
import os.path
# from machinevisiontoolbox import VideoCamera
# import time
# camera = VideoCamera(1)
# time.sleep(10)
# for i in range(10):
# image = camera.grab()
# time.sleep(0.1)
# camera.release()
# image.disp()
# from machinevisiontoolbox import *
from machinevisiontoolbox import Image
church = Image.Read("shark2.png")
print(church.metadata())
church.disp(block=True)
# exec(open(pathlib.Path(__file__).parent.parent.absolute() / "tests" / "test_processing.py").read()) # pylint: disable=exec-used