1
|
|
|
# Copyright 2014 Diamond Light Source Ltd. |
2
|
|
|
# |
3
|
|
|
# Licensed under the Apache License, Version 2.0 (the "License"); |
4
|
|
|
# you may not use this file except in compliance with the License. |
5
|
|
|
# You may obtain a copy of the License at |
6
|
|
|
# |
7
|
|
|
# http://www.apache.org/licenses/LICENSE-2.0 |
8
|
|
|
# |
9
|
|
|
# Unless required by applicable law or agreed to in writing, software |
10
|
|
|
# distributed under the License is distributed on an "AS IS" BASIS, |
11
|
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
12
|
|
|
# See the License for the specific language governing permissions and |
13
|
|
|
# limitations under the License. |
14
|
|
|
|
15
|
|
|
""" |
16
|
|
|
.. module:: rotate_90 |
17
|
|
|
:platform: Unix |
18
|
|
|
:synopsis: (Change this) A template to create a simple plugin that takes |
19
|
|
|
one dataset as input and returns a similar dataset as output. |
20
|
|
|
|
21
|
|
|
.. moduleauthor:: Jacob Williamson <[email protected]> |
22
|
|
|
""" |
23
|
|
|
from copy import deepcopy |
24
|
|
|
|
25
|
|
|
from savu.plugins.utils import register_plugin |
26
|
|
|
from savu.plugins.plugin import Plugin |
27
|
|
|
from savu.plugins.driver.cpu_plugin import CpuPlugin |
28
|
|
|
from savu.core.iterate_plugin_group_utils import enable_iterative_loop, \ |
29
|
|
|
check_if_end_plugin_in_iterate_group, setup_extra_plugin_data_padding |
30
|
|
|
from savu.data.data_structures.data_types.data_plus_darks_and_flats \ |
31
|
|
|
import ImageKey |
32
|
|
|
|
33
|
|
|
import numpy as np |
34
|
|
|
import h5py |
35
|
|
|
@register_plugin |
36
|
|
|
class Rotate90(Plugin, CpuPlugin): |
37
|
|
|
# Each class must inherit from the Plugin class and a driver |
38
|
|
|
|
39
|
|
|
def __init__(self): |
40
|
|
|
super(Rotate90, self).__init__("Rotate90") |
41
|
|
|
|
42
|
|
|
def nInput_datasets(self): |
43
|
|
|
return 1 |
44
|
|
|
|
45
|
|
|
|
46
|
|
|
def nOutput_datasets(self): |
47
|
|
|
return 1 |
48
|
|
|
|
49
|
|
|
|
50
|
|
|
def setup(self): |
51
|
|
|
|
52
|
|
|
# assumes 3D data and 2D frames |
53
|
|
|
|
54
|
|
|
if self.exp.meta_data.get("pre_run"): |
55
|
|
|
self.stats_obj.calc_stats = False |
56
|
|
|
|
57
|
|
|
in_dataset, out_dataset = self.get_datasets() |
58
|
|
|
pattern = self.parameters["pattern"] |
59
|
|
|
data_info = in_dataset[0].data_info |
60
|
|
|
core_dims = data_info["data_patterns"][pattern]["core_dims"] |
61
|
|
|
c0, c1 = core_dims[0], core_dims[1] # core dimensions |
62
|
|
|
s0 = data_info["data_patterns"][pattern]["slice_dims"][0] # slice dimension |
63
|
|
|
|
64
|
|
|
# swapping round core dimensions in the shape due to rotation |
65
|
|
|
new_shape = list(data_info["shape"]) |
66
|
|
|
new_shape[c0], new_shape[c1] = data_info["shape"][c1], data_info["shape"][c0] |
67
|
|
|
if self.exp.meta_data.get("pre_run"): |
68
|
|
|
new_shape[0] = in_dataset[0].data.image_key.shape[0] |
69
|
|
|
new_shape = tuple(new_shape) |
70
|
|
|
|
71
|
|
|
# swapping round core dimensions in axis labels |
72
|
|
|
new_axis_labels = deepcopy(data_info["axis_labels"]) |
73
|
|
|
new_axis_labels[c0], new_axis_labels[c1] = data_info["axis_labels"][c1], data_info["axis_labels"][c0] |
74
|
|
|
|
75
|
|
|
# swapping round core dimensions in data patterns |
76
|
|
|
new_data_patterns = deepcopy(data_info["data_patterns"]) |
77
|
|
|
for pattern in new_data_patterns: |
78
|
|
|
for dims in new_data_patterns[pattern]: |
79
|
|
|
if dims != "main_dir": # not sure what main_dir is ( = slice dim?) |
80
|
|
|
dims_list = list(new_data_patterns[pattern][dims]) |
81
|
|
|
for i, dim in enumerate(dims_list): |
82
|
|
|
if dim == c0: |
83
|
|
|
dims_list[i] = c1 |
84
|
|
|
elif dim == c1: |
85
|
|
|
dims_list[i] = c0 |
86
|
|
|
new_data_patterns[pattern][dims] = tuple(dims_list) |
87
|
|
|
|
88
|
|
|
dtype = in_dataset[0].dtype |
89
|
|
|
if dtype is None: |
90
|
|
|
dtype = in_dataset[0].data.data.dtype |
91
|
|
|
|
92
|
|
|
# creating output dataset with new axis, shape and data patterns to reflect rotated image |
93
|
|
|
if dtype: |
94
|
|
|
out_dataset[0].create_dataset(shape=new_shape, axis_labels=new_axis_labels, dtype=dtype) |
95
|
|
|
else: |
96
|
|
|
out_dataset[0].create_dataset(shape=new_shape, axis_labels=new_axis_labels) |
97
|
|
|
out_dataset[0].data_info.set("data_patterns", new_data_patterns) |
98
|
|
|
|
99
|
|
|
|
100
|
|
|
in_pData, out_pData = self.get_plugin_datasets() |
101
|
|
|
|
102
|
|
|
in_pData[0].plugin_data_setup(self.parameters['pattern'], 'single') |
103
|
|
|
out_pData[0].plugin_data_setup(self.parameters['pattern'], 'single') |
104
|
|
|
|
105
|
|
|
|
106
|
|
|
def pre_process(self): |
107
|
|
|
if self.exp.meta_data.get("pre_run"): |
108
|
|
|
in_dataset, out_dataset = self.get_datasets() |
109
|
|
|
dark = in_dataset[0].data.dark() |
110
|
|
|
flat = in_dataset[0].data.flat() |
111
|
|
|
if dark.size: |
112
|
|
|
in_dataset[0].data.update_dark(self.process_frames_3d(dark)) |
113
|
|
|
if flat.size: |
114
|
|
|
in_dataset[0].data.update_flat(self.process_frames_3d(flat)) |
115
|
|
|
out_dataset[0].data.image_key = in_dataset[0].data.image_key |
116
|
|
|
|
117
|
|
|
def process_frames(self, data): |
118
|
|
|
# assumes 2D frame |
119
|
|
|
if self.parameters["direction"] == "ACW": |
120
|
|
|
data[0] = np.rot90(data[0], axes=(0, 1)) |
121
|
|
|
elif self.parameters["direction"] == "CW": |
122
|
|
|
data[0] = np.rot90(data[0], axes=(1, 0)) |
123
|
|
|
return data[0] |
124
|
|
|
|
125
|
|
|
def process_frames_3d(self, data): |
126
|
|
|
# assumes 3D frame |
127
|
|
|
if self.parameters["direction"] == "ACW": |
128
|
|
|
data = np.rot90(data, axes=(1, 2)) |
129
|
|
|
elif self.parameters["direction"] == "CW": |
130
|
|
|
data = np.rot90(data, axes=(2, 1)) |
131
|
|
|
return data |
132
|
|
|
|
133
|
|
|
def post_process(self): |
134
|
|
|
if self.exp.meta_data.get("pre_run"): |
135
|
|
|
in_dataset, out_dataset = self.get_datasets() |
136
|
|
|
image_key = in_dataset[0].data.image_key |
137
|
|
|
|
138
|
|
|
dark = in_dataset[0].data.dark_updated |
139
|
|
|
flat = in_dataset[0].data.flat_updated |
140
|
|
|
|
141
|
|
|
new_image_key = np.array([0.] * len(image_key)) |
|
|
|
|
142
|
|
|
new_image_key[- len(dark):] = [2.] * len(dark) |
143
|
|
|
new_image_key[- len(dark) - len(flat): - len(dark)] = [1.] * len(flat) |
144
|
|
|
|
145
|
|
|
out_dataset[0].data[- len(dark):] = dark |
146
|
|
|
out_dataset[0].data[- len(dark) - len(flat): - len(dark)] = flat |
147
|
|
|
|
148
|
|
|
out_dataset[0].data.image_key = new_image_key |
149
|
|
|
|
150
|
|
|
out_dataset[0].data = ImageKey(out_dataset[0], new_image_key, 0) |
151
|
|
|
|