1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
|
# ----------------------------------------------------------------------------
# - Open3D: www.open3d.org -
# ----------------------------------------------------------------------------
# Copyright (c) 2018-2024 www.open3d.org
# SPDX-License-Identifier: MIT
# ----------------------------------------------------------------------------
import os
import configargparse
import open3d as o3d
class ConfigParser(configargparse.ArgParser):
def __init__(self):
super().__init__(default_config_files=[
os.path.join(os.path.dirname(__file__), 'default_config.yml')
],
conflict_handler='resolve')
# yapf:disable
# Default arguments
self.add(
'--name', type=str,
help='Name of the config for the offline reconstruction system.')
self.add(
'--fragment_size', type=int,
help='Number of RGBD frames to construct a fragment.')
self.add(
'--device', type=str,
help='Device to run the system.')
self.add(
'--engine', type=str,
choices=['tensor', 'legacy'],
help='Open3D engine to reconstruct.')
self.add(
'--multiprocessing', action='store_true',
help='Use multiprocessing in operations. Only available for the legacy engine.')
input_parser = self.add_argument_group('input')
input_parser.add(
'--path_dataset', type=str,
help='Path to the dataset folder. It should contain a folder with depth and a folder with color images.')
input_parser.add(
'--depth_folder', type=str,
help='Path that stores depth images.')
input_parser.add(
'--color_folder', type=str,
help='Path that stores color images.')
input_parser.add(
'--path_intrinsic', type=str,
help='Path to the intrinsic.json config file.'
'If the intrinsic matrix for color image is different,'
'specify it by --path_color_intrinsic.'
'By default PrimeSense intrinsics is used.')
input_parser.add(
'--path_color_intrinsic', type=str,
help='Path to the intrinsic.json config file.'
'If the intrinsic matrix for color image is different,'
'specify it by --path_color_intrinsic.'
'By default PrimeSense intrinsics is used.')
input_parser.add(
'--depth_min', type=float,
help='Min clipping distance (in meter) for input depth data.')
input_parser.add(
'--depth_max', type=float,
help='Max clipping distance (in meter) for input depth data.')
input_parser.add(
'--depth_scale', type=float,
help='Scale factor to convert raw input depth data to meters.')
input_parser.add('--fragment_size', type=int, help='Number of RGBD frames per fragment')
odometry_parser = self.add_argument_group('odometry')
odometry_parser.add(
'--odometry_method', type=str,
choices=['point2plane', 'intensity', 'hybrid', 'frame2model'],
help='Method used in pose estimation between RGBD images.'
'Frame2model only available for the tensor engine.')
odometry_parser.add(
'--odometry_loop_interval', type=int,
help='Intervals to check loop closures between RGBD images.')
odometry_parser.add(
'--odometry_loop_weight', type=float,
help='Weight of loop closure edges when optimizing pose graphs for odometry.')
odometry_parser.add(
'--odometry_distance_thr', type=float,
help='Default distance threshold to filter outliers in odometry correspondences.')
registration_parser = self.add_argument_group('registration')
registration_parser.add(
'--icp_method', type=str,
choices=['colored', 'point2point', 'point2plane', 'generalized'],
help='Method used in registration between fragment point clouds with a good initial pose estimate.'
'Generalized ICP only available for the tensor engine.')
registration_parser.add(
'--icp_voxelsize', type=float,
help='Voxel size used to down sample point cloud for fast/multiscale ICP.')
registration_parser.add(
'--icp_distance_thr', type=float,
help='Default distance threshold to filter outliers in ICP correspondences.')
registration_parser.add(
'--global_registration_method', type=str,
choices=['fgr', 'ransac'],
help='Method used in global registration of two fragment point clouds without an initial pose estimate.')
registration_parser.add(
'--registration_loop_weight', type=float,
help='Weight of loop closure edges when optimizing pose graphs for registration.')
integration_parser = self.add_argument_group('integration')
integration_parser.add(
'--integrate_color', action='store_true',
default=False, help='Volumetric integration mode.')
integration_parser.add(
'--voxel_size', type=float,
help='Voxel size in meter for volumetric integration.')
integration_parser.add(
'--trunc_voxel_multiplier', type=float,
help='Truncation distance multiplier in voxel size for signed distance. For instance, --trunc_voxel_multiplier=8 with --voxel_size=0.006(m) creates a truncation distance of 0.048(m).')
integration_parser.add(
'--est_point_count', type=int,
help='Estimated point cloud size for surface extraction.')
integration_parser.add(
'--block_count', type=int,
help='Pre-allocated voxel block count for volumetric integration.')
integration_parser.add(
'--surface_weight_thr', type=float,
help='Weight threshold to filter outliers during volumetric surface reconstruction.')
# yapf:enable
def get_config(self):
config = self.parse_args()
# Resolve conflicts
if config.engine == 'legacy':
if config.device.lower().startswith('cuda'):
print('Legacy engine only supports CPU.', 'Fallback to CPU.')
config.device = 'CPU:0'
if config.odometry_method == 'frame2model':
print('Legacy engine does not supports frame2model tracking.',
'Fallback to hybrid odometry.')
config.odometry_method = 'hybrid'
elif config.engine == 'tensor':
if config.icp_method == 'generalized':
print('Tensor engine does not support generalized ICP.',
'Fallback to colored ICP.')
config.icp_method = 'colored'
if config.multiprocessing:
print('Tensor engine does not support multiprocessing.',
'Disabled.')
config.multiprocessing = False
if (config.device.lower().startswith('cuda') and
(not o3d.core.cuda.is_available())):
print(
'Open3d not built with cuda support or no cuda device available. ',
'Fallback to CPU.')
config.device = 'CPU:0'
return config
if __name__ == '__main__':
# Priority: command line > custom config file > default config file
parser = ConfigParser()
parser.add(
'--config',
is_config_file=True,
help='YAML config file path. Please refer to default_config.yml as a '
'reference. It overrides the default config file, but will be '
'overridden by other command line inputs.')
config = parser.get_config()
print(config)
|
