-
Notifications
You must be signed in to change notification settings - Fork 1
/
img_utils.py
107 lines (100 loc) · 4.2 KB
/
img_utils.py
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
import numpy as np
from os import listdir
BASE_DIR_THEBE_TRAIN_SEISMIC = '/data/cvfault/thebeData/processedThebe/train/seismic'
BEATRICE_RAW_DATA = './dataset_raw/beatrice.npy'
THEBE_TRAIN_DATA = './dataset_raw/thebe_train.npy'
FAULT_DATA = '/data/cvfault/faultSegData/validation/seis/{}.dat'
# FAULT_SELECTED_DATA = [4,13,14,20,23,24,100,121,130,116,118,111,112,113,105,106,107,109,86,98,59,57,39,62,63,91,183,199,171,191,159,166]
# FAULT_SELECTED_DATA = [1,12,41,42,74,95,97,64,54,49,103,105,113,115,120,121,124,134,145,150,155,161,163,164,170,182,191,195,106,117,128,136,146,148,188,198,199]
# GSB_DATA = '/data/cvfault/gsbData/gsb_{}.npy'
# GSB_SUB_TRACE = ['crl_2600','crl_2700','crl_2800','crl_2900','crl_3000','inl_1791','inl_2011']
def normalize(img, min_val = 0, max_val = 1):
return (img - np.min(img)) / (np.max(img) - np.min(img)) * (max_val - min_val) + min_val
def rgb_to_gray(img):
if len(img.shape) == 3:
img = .03*img[:,:,0]+0.59*img[:,:,1]+0.11*img[:,:,2]
return img
def dataloader(trace_name, seismic):
result = []
if trace_name == 'thebe' and seismic:
files = [file for file in listdir(BASE_DIR_THEBE_TRAIN_SEISMIC) if not file.startswith('.')]
files = files[:10000]
for file in files:
result.append(np.load(BASE_DIR_THEBE_TRAIN_SEISMIC + '/{}'.format(file)))
return files, np.asarray(result)
def full_trace(trace_name):
rtn = None
if trace_name == 'beatrice':
print("loading data from beatrice")
rtn = np.rot90(np.load(BEATRICE_RAW_DATA), axes = (1,2))
print("load success, data shape is {}".format(rtn.shape))
return rtn
if trace_name == 'thebe_train':
print("loading data from thebe train")
rtn = np.rot90(np.load(THEBE_TRAIN_DATA), 3, axes = (1,2))
print("load success, data shape is {}".format(rtn.shape))
return rtn
if trace_name == 'thebe_val':
print("loading data from thebe val")
rtn = np.rot90(np.load(THEBE_VAL_DATA), 3, axes = (1,2))
print("load success, data shape is {}".format(rtn.shape))
return rtn
if trace_name == 'thebe_test':
print("loading data from thebe test")
rtn = np.rot90(np.load(THEBE_TEST_DATA), 3, axes = (1,2))
print("load success, data shape is {}".format(rtn.shape))
return rtn
if trace_name == 'faultseg':
print("loading data from thebe test (only selected)")
rtn = None
for num in range(20):
dat = np.fromfile(FAULT_DATA.format(num),dtype=np.single).reshape(128,128,128)
dat = np.rot90(dat, axes = (1,2))
if rtn is None:
rtn = dat
else:
rtn = np.concatenate((rtn,dat), axis=0)
# rtn = np.array(rtn)
print("load success, data shape is {}".format(rtn.shape))
return rtn
if trace_name == 'gsb':
print("loading data from gsb")
rtn=[]
for file_name in GSB_SUB_TRACE:
gsb_np = np.load(GSB_DATA.format(file_name))[0]
print(gsb_np.shape)
max_x,max_y = gsb_np.shape
if (max_x > max_y):
gsb_np = np.rot90(gsb_np)
if gsb_np.shape[1] > 484:
gsb_np = gsb_np[:,:484]
rtn.append(gsb_np)
rtn = np.array(rtn)
print("load success, data shape is {}".format(rtn.shape))
return rtn
print("wrong param, load failed!")
return rtn
def gaussian_noise(img, mean, sigma):
'''
此函数用将产生的高斯噪声加到图片上
传入:
img : 原图
mean : 均值
sigma : 标准差
返回:
gaussian_out : 噪声处理后的图片
noise : 对应的噪声
'''
# 将图片灰度标准化
img = img / 255
# 产生高斯 noise
noise = np.random.normal(mean, sigma, img.shape)
# 将噪声和图片叠加
gaussian_out = img + noise
# 将超过 1 的置 1，低于 0 的置 0
gaussian_out = np.clip(gaussian_out, 0, 1)
# 将图片灰度范围的恢复为 0-255
gaussian_out = np.uint8(gaussian_out*255)
# 将噪声范围搞为 0-255
# noise = np.uint8(noise*255)
return gaussian_out, noise # 这里也会返回噪声，注意返回值