-
Notifications
You must be signed in to change notification settings - Fork 4
/
Copy pathUtilsCMS.py
156 lines (143 loc) · 6.78 KB
/
UtilsCMS.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
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
# -*- coding:utf-8 -*-
# Author：Mingshuo Cai
# Create_time：2023-08-01
# Updata_time：2024-03-15
# Usage：Implementation of the GuidedPGC proposed in MLUDA.
import math
import pandas as pd
import torch
import torch.nn as nn
import torch.nn.functional as F
from torch.autograd import Variable
import mmd
import numpy as np
from sklearn import metrics
import time
import utils
from torch.utils.data import TensorDataset, DataLoader
from contrastive_loss import SupConLoss
# from config_Houston import *
from sklearn import svm
import cv2
import hdf5storage
import matplotlib.pyplot as plt
plt.rcParams['font.sans-serif'] = ['SimHei']
# 自己加入的库
from sklearn.decomposition import PCA
from skimage import exposure
import argparse
import cv2
def PlotColor(label):
"将Label以热力图的形式进行可视化"
plt.imshow(label, cmap='viridis', interpolation='nearest')
plt.colorbar() # 添加颜色条
plt.title('Source')
plt.show()
def pca(data,n):
"进行pca过程的代码"
pca = PCA(n_components=n) # 定义pca的方法
height, width, channels = data.shape
data = data.reshape(-1,channels)
data_PCA = pca.fit_transform(data).reshape(height,width,n)
Score = pca.explained_variance_ratio_
# 检验三个主成分是否可以代表
# print("前三个主成分的贡献程度:")
# print(Score)
# 在每个通道上进行归一化操作 有没有可能这里的归一化会把东西搞坏
min_value = np.min(data_PCA, axis=(0, 1)) # 在前两个维度上找最小值
max_value = np.max(data_PCA, axis=(0, 1)) # 在前两个维度上找最大值
data_PCA = (data_PCA - min_value) / (max_value - min_value)
return data_PCA,Score
def gammaCorrect(SourcePic,TargetPic):
"""
进行gamma校准的代码
SourcePic:一个传入源域的三维的图像
TargetPic:一个传入目标的三维的图像
"""
# 分别计算源域和目标域的亮度
# Sourcebrightness1 = 0.299 * SourcePic[:,:,0] + 0.587 * SourcePic[:,:,1] + 0.114 * SourcePic[:,:,2]
# Targetbrightness1 = 0.299 * TargetPic[:,:,0] + 0.587 * TargetPic[:,:,1] + 0.114 * TargetPic[:,:,2]
Sourcebrightness1 = np.mean(SourcePic)
Targetbrightness1 = np.mean(TargetPic)
# 计算需要应用的Gamma值，以使两个图像的亮度一致
average_brightness_Source = np.mean(Sourcebrightness1)
average_brightness_Target = np.mean(Targetbrightness1)
gamma_value = abs(average_brightness_Source / average_brightness_Target )
# 对源域和目标域的图像分别应用gamma校准
Source_corrected = np.power(SourcePic,gamma_value)
Target_corrected = TargetPic
Target_corrected = np.clip(Target_corrected, 0, 1)
# Sourcebrightness2 = (0.299 * Source_corrected[:,:,0]
# + 0.587 * Source_corrected[:,:,1] + 0.114 * Source_corrected[:,:,2])
# Targetbrightness2 = (0.299 * Target_corrected[:,:,0]
# + 0.587 * Target_corrected[:,:,1] + 0.114 * Target_corrected[:,:,2])
Sourcebrightness2 = np.mean(Source_corrected)
Targetbrightness2 = np.mean(Target_corrected)
print("gamma:"+str(gamma_value))
print(Sourcebrightness1.mean())
print(Targetbrightness1.mean())
print(Sourcebrightness2.mean())
print(Targetbrightness2.mean())
Brightness = [Sourcebrightness1.sum(),Targetbrightness1.sum(),
Sourcebrightness2.sum(),Targetbrightness2.sum()]
return Source_corrected,Target_corrected,Brightness
def colorAdaption(Source_image,Target_image):
# 进行颜色直方图匹配的代码
Source_image_Matched = exposure.match_histograms(Source_image, Target_image)
return Source_image_Matched,Target_image
def GuideFilter(Pic_MChannels, Pic_3Channels,r):
# 进行导向滤波的代码
Pic_3Channels = Pic_3Channels.astype(np.float32)
Pic_MChannels = Pic_MChannels.astype(np.float32)
# 执行导向滤波
print("1, " + str(r))
radius = 1 # 滤波半径
epsilon = r # 正则化参数
filtered_image = cv2.ximgproc.guidedFilter(Pic_3Channels,Pic_MChannels, radius, epsilon)
return filtered_image
def TotalAdaption(Source_Pic, Target_Pic, pca_n,r):
"""
进行图像级域迁移的代码
"""
Source_data, Source_pca_score = pca(Source_Pic, pca_n)
Target_data, Target_pca_score = pca(Target_Pic, pca_n)
# Pca之前的操作都是没有问题的
Source_data_gamma, Target_data_gamma, Brightness = gammaCorrect(Source_data, Target_data) # 经过gamma校准
print(Source_data_gamma.shape,Target_data_gamma.shape)
Source_data_Color, Target_data_Color = colorAdaption(Source_data_gamma, Target_data_gamma) # 经过颜色直方图匹配
print(Source_data_Color.shape,Target_data_Color.shape)
# print("经过gamma校准和颜色直方图匹配之后的效果展示：")
# fig, axes = plt.subplots(3, 2, figsize=(90, 10))
# axes[0, 0].imshow(Source_data)
# axes[0, 0].set_title('源域经过PCA之后的三通道,亮度为' + str(Brightness[0])) # 设置子图标题
# axes[0, 1].imshow(Target_data)
# axes[0, 1].set_title('目标域经过PCA之后的三通道,亮度为' + str(Brightness[1])) # 设置子图标题
# axes[1, 0].imshow(Source_data_gamma)
# axes[1, 0].set_title('源域经过PCA，gamma校准之后的三通道,亮度为' + str(Brightness[2])) # 设置子图标题
# axes[1, 1].imshow(Target_data_gamma)
# axes[1, 1].set_title('目标域经过PCA,gamma校准之后的三通道,亮度为' + str(Brightness[3])) # 设置子图标题
# axes[2, 0].imshow(Source_data_Color)
# axes[2, 0].set_title('源域经过颜色直方图匹配之后的三通道:') # 设置子图标题
# axes[2, 1].imshow(Target_data_Color)
# axes[2, 1].set_title('目标域经过颜色直方图之后的三通道:') # 设置子图标题
# plt.savefig('PIC/Gamma.png', dpi=300)
Final_Source = GuideFilter(Source_Pic, Source_data_Color,r)
Final_Target = GuideFilter(Target_Pic, Target_data_Color,r)
return Final_Source, Final_Target
def ILDA(data_s,data_t,pca_n,r):
# if param is None:
# # Pavia数据集
# # data_path_s = './datasets/Pavia/paviaU.mat'
# # label_path_s = './datasets/Pavia/paviaU_gt_7.mat'
# # data_path_t = './datasets/Pavia/pavia.mat'
# # label_path_t = './datasets/Pavia/pavia_gt_7.mat'
# # Houston数据集
# data_path_s = './datasets/Houston/Houston13.mat'
# label_path_s = './datasets/Houston/Houston13_7gt.mat'
# data_path_t = './datasets/Houston/Houston18.mat'
# label_path_t = './datasets/Houston/Houston18_7gt.mat'
# data_s,label_s = load_data_houston(data_path_s,label_path_s)
# data_t,label_t = load_data_houston(data_path_t,label_path_t)
# 图像级迁移的执行部分
ILDA_S,ILDA_T= TotalAdaption(data_s,data_t,pca_n,r)
return ILDA_S,ILDA_T