-
Notifications
You must be signed in to change notification settings - Fork 1
/
data_input.py
136 lines (120 loc) · 4.41 KB
/
data_input.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
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
# coding:utf-8
"""
总共三个场景，每个场景有80个训练数据，20个测试数据
训练时随机从每个场景中取50个数据
测试时随机从每个场景中取10个数据
ACB-1,AnFpark-2,PDFpark-3
"""
import numpy as np
import random
import gabor_gist
import pylab as pl
from PIL import Image
def getTrainSample():
# 获得每个场景的要取的图片
ACB_index = []
AnFpark_index = []
PDFpark_index = []
for i in range(80):
ACB_index.append(i)
AnFpark_index.append(i)
PDFpark_index.append(i)
random.shuffle(ACB_index)
random.shuffle(AnFpark_index)
random.shuffle(PDFpark_index)
# 读取图片
img_array = []
img_labels = []
for i in range(50):
image = Image.open('TrainData/ACB/' + 'ACB_' + str(ACB_index[i])+'.jpeg')
img_array.append(image)
img_labels.append(1)
image = Image.open('TrainData/AnFpark/' + 'AnFpark_' + str(AnFpark_index[i])+'.jpeg')
img_array.append(image)
img_labels.append(2)
image = Image.open('TrainData/FDFpark/' + 'FDFpark_' + str(PDFpark_index[i])+'.jpeg')
img_array.append(image)
img_labels.append(3)
return img_array,img_labels
def getTestSample():
# 获得每个场景的要取的图片
ACB_index = []
AnFpark_index = []
PDFpark_index = []
for i in range(80,100):
ACB_index.append(i)
AnFpark_index.append(i)
PDFpark_index.append(i)
random.shuffle(ACB_index)
random.shuffle(AnFpark_index)
random.shuffle(PDFpark_index)
# 读取图片
img_array = []
img_labels = []
for i in range(10):
image = Image.open('TestData/ACB/' + 'ACB_' + str(ACB_index[i])+'.jpeg')
img_array.append(image)
img_labels.append(1)
image = Image.open('TestData/AnFpark/' + 'AnFpark_' + str(AnFpark_index[i])+'.jpeg')
img_array.append(image)
img_labels.append(2)
image = Image.open('TestData/FDFpark/' + 'FDFpark_' + str(PDFpark_index[i])+'.jpeg')
img_array.append(image)
img_labels.append(3)
return img_array,img_labels
def getTrainResult():
features = np.load('TrainResult/features.npy')
img_labels = np.load('TrainResult/img_labels.npy')
pca_array = np.load('TrainResult/pca_array.npy')
return features,img_labels,pca_array
class InputClass:
features, img_labels, pca_array = getTrainResult()
train_images,train_labels = getTrainSample()
test_images,test_labels = getTestSample()
temp_data = pca_array
temp_label = img_labels
'''
返回的图片必须是二维数组，每一行是一个80维的gist特征，一次返回num个gist特征
返回的标签也必须是一个二维数组，每一行是一个三维的标签，除了对应的真实值为1，剩余两个都是0，一次返回num个标签组
'''
def next_batch(self,num):
r_data = []
r_label = []
# 获得每个场景的要取的图片
data_index = []
if len(self.temp_data)<num:
self.temp_data = self.pca_array
self.temp_label = self.img_labels
for i in range(len(self.temp_data)):
data_index.append(i)
random.shuffle(data_index)
for i in range(num):
r_data.append(self.temp_data[data_index[i]])
label = [0, 0, 0]
label[self.temp_label[data_index[i]] - 1] = 1
r_label.append(label)
next_data = []
next_label = []
for i in range(num,len(self.temp_data)):
next_data.append(self.temp_data[data_index[i]])
next_label.append(self.temp_label[data_index[i]])
self.temp_data = np.array(next_data)
self.temp_label = np.array(next_label)
return np.array(r_data), np.array(r_label)
def test_data(self):
# 求每个测试数据的标签
scale = [5, 8, 11, 14]
test_gists = []
test_label_arr = []
for i in range(len(self.test_images)):
# 转灰度
gray_img = self.test_images[i].convert('L')
# gist特征
gist = gabor_gist.getGist(gray_img, 4, 8, scale)
# 降维
pca_gist = np.dot(np.array(gist), np.transpose(self.features))
test_gists.append(pca_gist)
label = [0, 0, 0]
label[self.test_labels[i] - 1] = 1
test_label_arr.append(label)
return np.array(test_gists),np.array(test_label_arr)