# This file is part of Sympathy for Data.
# Copyright (c) 2017, Combine Control Systems AB
#
# Sympathy for Data is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, version 3 of the License.
#
# Sympathy for Data is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Sympathy for Data. If not, see <http://www.gnu.org/licenses/>.
from sympathy.api import node
from sympathy.api.nodeconfig import Port, Ports, Tag, Tags
import numpy as np
from sylib.imageprocessing.image import ImagePort
from sylib.imageprocessing.image import File as ImageFile
from sylib.imageprocessing.algorithm_selector import (
ImageFiltering_abstract)
[docs]class ImageToList(ImageFiltering_abstract, node.Node):
name = 'Image to List'
author = 'Mathias Broxvall'
version = '0.1'
icon = 'image_list.svg'
description = ('Generates a list of images based on algorithms operating '
'on one image')
nodeid = 'syip.image2list'
tags = Tags(Tag.ImageProcessing.Layers)
def alg_from_labels(im, results, par):
if not np.issubdtype(im.dtype, np.integer):
# Early return, input is not a labeled image
return
if len(im.shape) == 3:
im = im[:, :, 0]
for value in range(np.min(im), np.max(im)+1):
mask = im == value
if value == 0 and not par['do background'].value:
pass
elif np.sum(mask) > 0:
new_image = ImageFile()
new_image.set_image(mask)
results.append(new_image)
def alg_from_channels(im, results, par):
if len(im.shape) < 3:
im = im.reshape(im.shape + (1,))
channels = im.shape[2]
for channel in range(channels):
new_image = ImageFile()
new_image.set_image(im[:, :, channel])
results.append(new_image)
algorithms = {
'from labels': {
'description': ('Generates an image mask selecting each label in '
'an image once'),
'do background': 'If true include the background object (ID=0)',
'multi_chromatic': False,
'algorithm': alg_from_labels},
'from channels': {
'description': ('Generates a list of grayscale images from each '
'channel in input image'),
'multi_chromatic': True,
'algorithm': alg_from_channels}}
options_list = ['do background']
options_types = {'do background': bool}
options_default = {'do background': True}
parameters = node.parameters()
parameters.set_string('algorithm', value=next(iter(algorithms)),
description='', label='Algorithm')
ImageFiltering_abstract.generate_parameters(parameters, options_types,
options_default)
inputs = Ports([
ImagePort('source image', name='source')])
outputs = Ports([
Port.Custom('[image]', 'Resulting list of images', name='results')])
__doc__ = ImageFiltering_abstract.generate_docstring(
description, algorithms, options_list, inputs, outputs)
def execute(self, node_context):
image = node_context.input['source'].get_image()
params = node_context.parameters
alg_name = params['algorithm'].value
results = node_context.output['results']
alg = self.algorithms[alg_name]['algorithm']
alg(image, results, params)
[docs]class ListToImage(ImageFiltering_abstract, node.Node):
name = 'List to Image'
author = 'Mathias Broxvall'
version = '0.1'
icon = 'image_list2image.svg'
description = 'Generates an image based on a list of images'
nodeid = 'syip.list2image'
tags = Tags(Tag.ImageProcessing.Layers)
def alg_concatenate(images, par):
max_x, max_y = 0, 0
channels = 0
if len(images) == 0:
return np.zeros((1, 1, 1))
dtype = images[0].get_image().dtype
for im_obj in images:
im = im_obj.get_image()
max_x = max(max_x, im.shape[1])
max_y = max(max_y, im.shape[0])
if len(im.shape) > 2:
channels += im.shape[2]
else:
channels += 1
result = np.zeros((max_y, max_x, channels), dtype=dtype)
ch = 0
for im_obj in images:
im = im_obj.get_image()
if len(im.shape) < 3:
im = im.reshape(im.shape + (1,))
result[:im.shape[0], :im.shape[1], ch:ch+im.shape[2]] = im
ch += im.shape[2]
return result
algorithms = {
'concatenate channels': {
'description': ('Generates an image by concatenating all channels '
'in the input images. Resulting datatype given by '
'first image in the list'),
'multi_chromatic': True,
'algorithm': alg_concatenate}}
options_list = ['do background']
options_types = {'do background': bool}
options_default = {'do background': True}
parameters = node.parameters()
parameters.set_string('algorithm', value=next(iter(algorithms)),
description='', label='Algorithm')
ImageFiltering_abstract.generate_parameters(parameters, options_types,
options_default)
inputs = Ports([
Port.Custom('[image]', 'Input list of images', name='inputs')
])
outputs = Ports([
ImagePort('result image', name='result'),
])
__doc__ = ImageFiltering_abstract.generate_docstring(
description,
algorithms, options_list, inputs, outputs)
def execute(self, node_context):
params = node_context.parameters
output = node_context.output['result']
inputs = node_context.input['inputs']
alg_name = params['algorithm'].value
alg = self.algorithms[alg_name]['algorithm']
im = alg(inputs, params)
output.set_image(im)
