deformernet_DNN/process_data.py at master · Kuntz-Lab/deformernet_DNN · GitHub

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
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
import h5py
import numpy as np
import pickle


import open3d
import os
import numpy as np
import pickle
import timeit
from farthest_point_sampling import *
# from pointnet2_ops_lib.pointnet2_ops.pointnet2_utils import furthest_point_sample, gather_operation

# resample = lambda points, n: gather_operation(points.transpose(1 , 2).contiguous(), furthest_point_sample(points, n))

data_recording_path = "/home/baothach/shape_servo_data/keypoints/combined_w_shape_servo/batch3_original_partial_pc/processed_2"
data_processed_path = "/home/baothach/shape_servo_data/keypoints/combined_w_shape_servo/batch3_original_partial_pc/processed_2"
# i = 1001
start_time = timeit.default_timer()

start_index = 0
max_len_data = 13000

for i in range(start_index, max_len_data):
    if i % 50 == 0:
        print("current count:", i, " , time passed:", timeit.default_timer() - start_time)

    file_name = os.path.join(data_recording_path, "processed sample " + str(i) + ".pickle")
    with open(file_name, 'rb') as handle:
        data = pickle.load(handle)

    # pc = data["point clouds"][0]
    # pc_goal = data["point clouds"][1]


    max_x = 0.4
    max_y = 0.4
    max_z = 0.2

    delta_x = np.random.uniform(low = -max_x, high = max_x)
    delta_y = np.random.uniform(low = -max_y, high = max_y)
    delta_z = np.random.uniform(low = 0, high = max_z)

    augmented_pc = data["point clouds"][0]
    # augmented_pc[0] += delta_x
    # augmented_pc[1] += delta_y
    # augmented_pc[2] += delta_z
    augmented_pc_goal = data["point clouds"][1]
    augmented_pc_goal[0] += delta_x
    augmented_pc_goal[1] += delta_y
    augmented_pc_goal[2] += delta_z

    positions = data["positions"]
    positions[0] -= delta_x
    positions[1] -= delta_y
    positions[2] += delta_z


    pcs = (augmented_pc, augmented_pc_goal)
    augmented_data = {"point clouds": pcs, "positions": positions, "grasp_pose": data["grasp_pose"]}
    with open(os.path.join(data_processed_path, "processed sample " + str(i+max_len_data) + ".pickle"), 'wb') as handle:
        pickle.dump(augmented_data, handle, protocol=pickle.HIGHEST_PROTOCOL)


# with open('/home/baothach/shape_servo_data/batch_3b(using_camera)', 'rb') as handle:
#     data1 = pickle.load(handle)


# final_point_clouds = data1["point clouds"]
# final_desired_positions = data1["positions"]
# count = 0
# modified_pcs = []
# for (pc, pc_goal) in final_point_clouds:
#     count += 1
#     print(count)
#     # if count == 2:
#     #     break

#     pcd = open3d.geometry.PointCloud()

#     pcd.points = open3d.utility.Vector3dVector(np.array(pc))
#     pcd.estimate_normals(
#         search_param=open3d.geometry.KDTreeSearchParamHybrid(radius=1,
#                                                             max_nn=10))

#     radii = [0.005, 0.01, 0.02, 0.04]
#     # radii =  [0.1, 0.2, 0.4, 0.8]
#     rec_mesh = open3d.geometry.TriangleMesh.create_from_point_cloud_ball_pivoting(
#         pcd, open3d.utility.DoubleVector(radii))

#     rec_mesh.compute_vertex_normals()
#     pcd = rec_mesh.sample_points_uniformly(number_of_points=3000)

#     pcd_2 = open3d.geometry.PointCloud()

#     pcd_2.points = open3d.utility.Vector3dVector(np.array(pc_goal))
#     pcd_2.estimate_normals(
#         search_param=open3d.geometry.KDTreeSearchParamHybrid(radius=1,
#                                                             max_nn=10))

#     radii = [0.005, 0.01, 0.02, 0.04]
#     # radii =  [0.1, 0.2, 0.4, 0.8]
#     rec_mesh = open3d.geometry.TriangleMesh.create_from_point_cloud_ball_pivoting(
#         pcd_2, open3d.utility.DoubleVector(radii))

#     rec_mesh.compute_vertex_normals()
#     pcd_2 = rec_mesh.sample_points_uniformly(number_of_points=3000)
#     modified_pcs.append((np.asarray(pcd.points), np.asarray(pcd_2.points)))


# data = {"point clouds": modified_pcs, "positions": final_desired_positions}
# with open('/home/baothach/shape_servo_data/batch_3b(using_camera)_modified', 'wb') as handle:
#     pickle.dump(data, handle, protocol=pickle.HIGHEST_PROTOCOL)


# f = h5py.File("/home/baothach/shape_servo_data/multi_grasps/batch_1")
# total_grasp = f['cur_grasp_id'][()] + 1
# print("total:", total_grasp)
# # grasp_poses = []
# # point_clouds = []
# # positions = []
# datas = []
# count = 0


# for i in range(total_grasp):
#     # pc_pairs = f['point clouds '+ str(i)][()]
#     count += 1
#     print(count)
#     for k, pc_pair in enumerate(f['point clouds '+ str(i)][()]):
#         position = f['positions '+ str(i)][()][k]
#         grasp_pose = f['manipulation pose '+ str(i)][()]
#         modified_pc = (np.swapaxes(pc_pair[0],0,1), np.swapaxes(pc_pair[1],0,1))
#         data = {"grasp pose": grasp_pose, "point clouds": modified_pc, "positions": position}
#         datas.append(data)

# with open('/home/baothach/shape_servo_data/multi_grasps/batch1_processed', 'wb') as handle:
#     pickle.dump(datas, handle, protocol=pickle.HIGHEST_PROTOCOL)