-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathPCA_SVM_GENDER.py
140 lines (126 loc) · 5.66 KB
/
PCA_SVM_GENDER.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
import numpy as np
import random, os, cv2
from sklearn.decomposition import PCA
import time
import pywt
from sklearn.svm import SVC
import joblib
import load_lable
from capture_Face import captureFace
import matplotlib.pyplot as plt
def loadImageSet(folder='.\wiki_crop'): #加载图像集，随机选择sampleCount张图片用于训练
trainData = [];
testData = [];
gender_Train=[];
gender_Test=[]
age_full,gender_full,path_full,sample_size=load_lable.load_lable(35000)
address1st=[]
for path in path_full :
path_cata=path[0:2]
path_filename=path[3:]
address1st.append(os.path.join(folder, path_cata,path_filename))
address=[]
age_2rd=[]
gender_2rd=[]
for iaddress in range(len(address1st)):
if os.path.exists(address1st[iaddress]):
address.append(address1st[iaddress])
age_2rd.append(age_full[iaddress])
gender_2rd.append(gender_full[iaddress])
else:
continue
print('-------address_validiation completed----------'+str(len(address)))
data = [cv2.imread(d, 0) for d in address]
age_full_renew=[]
gender_full_renew=[]
data_pwt = []
for m in range(len(address)):
tempimg,face_lenth,img_full=captureFace(data[m])
if face_lenth>0:
item=cv2.resize(tempimg,(200,200))
# cv2.imshow('frame2Q', item)
# cv2.waitKey(0)
# cv2.destroyAllWindows()
xTrain_pwt = pywt.wavedec2(item, 'sym2', mode='symmetric', level=4, axes=(-2, -1))
data_pwt.append(xTrain_pwt[0])
age_full_renew.append(age_2rd[m])
gender_full_renew.append(gender_2rd[m])
else:
continue
simple_size_renew=len(data_pwt)
print('-------------face_cut completed--------------'+str(simple_size_renew))
sampleCount=int(simple_size_renew*0.7)
sample = random.sample(range(simple_size_renew), sampleCount) # random.sample()可以从指定的序列中，随机的截取指定长度的片断，不作原地修改
trainData.extend([data_pwt[i].ravel() for i in range(simple_size_renew) if i in sample]) # ravel将多维数组降位一维####40*5*112*92
testData.extend([data_pwt[i].ravel() for i in range(simple_size_renew) if i not in sample]) # 40*5*112*92
gender_Train.extend(gender_full_renew[i] for i in range(simple_size_renew) if i in sample)
gender_Test.extend(gender_full_renew[i] for i in range(simple_size_renew) if i not in sample)
return np.array(trainData), np.array(gender_Train), np.array(testData), np.array(gender_Test)
def main():
#loadImageSet()
t_start_load = time.process_time()
xTrain_, yTrain, xTest_, yTest = loadImageSet()# 200*10304
t_end_load = time.process_time()
t_diff_load = t_end_load - t_start_load
print('loadimageset cpmpleted---------------------------{:.5f} s'.format(t_diff_load))
t_start_pca = time.process_time()
pca1=PCA(n_components=0.9)
#hyp.plot(xTrain_,'o')
xTrain_pca=pca1.fit_transform(xTrain_) # 把原始训练集映射到主成分组成的子空间中
xTest_pca=pca1.transform(xTest_)# 把原始测试集映射到主成分组成的子空间中
t_end_pca = time.process_time()
t_diff_pca = t_end_pca - t_start_pca
print('PCA completed---------------------------------{:.5f} s'.format(t_diff_pca))
joblib.dump(pca1, 'pca_wiki_model.m')
# hyp.plot(xTrain_pca, 'o', n_clusters=40)
clf=SVC(C=1000.0, cache_size=2000, class_weight='balanced', coef0=0.0,
decision_function_shape='ovr', degree=3, gamma=0.001, kernel='poly',
probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)
#hyp.plot(xTrain_pca, 'o')
t_start_SVM = time.process_time()
clf.fit(xTrain_pca,yTrain)
t_end_SVM = time.process_time()
t_diff_SVM = t_end_SVM - t_start_SVM
print('SVM completed---------------------------------{:.5f} s'.format(t_diff_SVM))
predict=clf.predict(xTest_pca)
# hyp.plot(predict.reshape(200,1),'o', n_clusters=40)
# hyp.plot(yTest.reshape(200,1), 'o', n_clusters=40)
print(clf.score(xTest_pca,yTest))
print(u'支持向量机识别率: %.2f%%' % ((predict == np.array(yTest)).mean()*100) )
print('------------------------')
joblib.dump(clf, 'svm_gender_model.m')
if __name__ == '__main__':
# main()
pca_m = joblib.load('pca_wiki_model.m')
svm_m = joblib.load('svm_gender_model.m')
cap = cv2.VideoCapture(0)
while(True):
ret,frame = cap.read()
frame= cv2.flip(frame, 1, dst=None)
tempimg, face_lenth,img_cap = captureFace(frame)
if face_lenth > 0:
item = cv2.resize(tempimg, (200, 200))
item = cv2.cvtColor(item, cv2.COLOR_RGB2GRAY)
# cv2.imshow('frame2Q', item)
# cv2.waitKey(0)
# cv2.destroyAllWindows()
xTrain_pwtpp = pywt.wavedec2(item, 'sym2', mode='symmetric', level=4, axes=(-2, -1))
xtrain_data = xTrain_pwtpp[0]
predictdata = []
predictdata.extend(xtrain_data.ravel())
test_pac = pca_m.transform(np.array(predictdata).reshape(1,-1))
predict2 = svm_m.predict(test_pac)
outcome=''
if(predict2== 0):
outcome='Female'
elif(predict2== 1):
outcome='Male'
cv2.putText(frame,outcome , (5, 50), cv2.FONT_HERSHEY_SIMPLEX, 2,(0, 255, 255),2)
cv2.imshow('frame2Q', frame)
else:
continue
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cap.release()
cv2.destroyAllWindows()