-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathface_recog_attend.py
206 lines (183 loc) · 8.86 KB
/
face_recog_attend.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
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
197
198
199
200
201
202
203
204
205
206
# -*- coding: utf-8 -*-
"""
Created on Sat Jun 30 18:11:37 2018
@author: Welkin
"""
import cv2
from align_custom import AlignCustom
from face_feature import FaceFeature
from mtcnn_detect import MTCNNDetect
from tf_graph import FaceRecGraph
import argparse
import json
import sys
import numpy as np
import pandas as pd
import time
import dlib
from collections import OrderedDict
from scipy.spatial import distance as dist
eye_ar_thresh = 0.3
eye_ar_frames = 3
FACIAL_LANDMARK_INDXS = OrderedDict([("mouth",(48,68)),("left_eyebrow",(22,27)),("right_eyebrow",(17,22)),("left_eye",(42,48)),('right_eye',(36,42)),('jaw',(0,17)),("nose",(27,36))])
(lStart,lEnd) = FACIAL_LANDMARK_INDXS['left_eye']
(rStart,rEnd) = FACIAL_LANDMARK_INDXS['right_eye']
start_time = time.time()
attendance = pd.read_csv(r'.\Attendance.csv',delimiter=',')
daily_attendance = pd.DataFrame(columns = ['Name','Attendance'],index=range(60))
daily_attendance['Name'][:2]=['Akash Yadav','Pawan Jha']
daily_attendance['Attendance'][:60]=[0]*60
path = r'.\shape_predictor_68_face_landmarks.dat'
predictor = dlib.shape_predictor(path)
def main(args):
mode = args.mode
#print(mode)
if mode == "camera":
face_recog()
elif mode == 'input':
new_data_entry()
else:
raise ValueError("Unimplemented Mode")
def face_recog():
print("Camera is On.\nFace Recognition in process...")
#for capturing Video from camera
cap = cv2.VideoCapture(0)#0 is for idefinite time
counter = 0
total = 0
while True:
ret,frame = cap.read()
#call detect face function it will return landmarks on face and bounding box
rects,landmark = face_detect.detect_face(frame,80)#min face size set to 100x100
if rects:
temp_rects = dlib.rectangle(rects[0][0],rects[0][1],rects[0][2],rects[0][3])
shape = predictor(frame,temp_rects)
coords = np.zeros((shape.num_parts,2),dtype="int")
for i in range(shape.num_parts):
coords[i] = (shape.part(i).x,shape.part(i).y)
shape = coords
leftEye = shape[lStart:lEnd]
rightEye = shape[rStart:rEnd]
leftEAR = e_a_r(leftEye)
rightEAR = e_a_r(rightEye)
#average eye aspect ratio of both eyes
EAR = (leftEAR + rightEAR)/(2.0)
if EAR < eye_ar_thresh:
counter += 1
else:
if counter >= eye_ar_frames:
total += 1
counter = 0
aligns = []#this list conatin matrix of aligned images
positions = []#this list contain position of user's face :left ,right,center
for (i,rect) in enumerate(rects):#rects contain list of bounding box(x1,y1,w,h)
aligned_face,face_pos = aligner.align(160,frame,landmark[i])
if len(aligned_face) == 160 and len(aligned_face[0]) == 160:
aligns.append(aligned_face)
positions.append(face_pos)
else:
print('Alignment of face is failed')
if len(aligns)>0:
#feed align or preprocessed image to function it will return 128D vector which contain unique feature
#of faces and also called generate embedding function take input as image return 128D vector features
feat_atrb = extract_feature.get_features(aligns)
#this function return name and percentage of accuracy with image
detected_face = findPeople(feat_atrb,positions)
count = 0#this is for counting attendance of each student
for i,rect in enumerate(rects):
count = 0
#this will draw bounding box on detected face
cv2.rectangle(frame,(rect[0],rect[1]),(rect[0]+rect[2],rect[1]+rect[3]),(0,255,0),)
cv2.putText(frame,detected_face[i][0]+"-"+str(detected_face[i][1])+"%",(rect[0],rect[1]),cv2.FONT_HERSHEY_SIMPLEX,1,(255,255,255),1,cv2.LINE_AA)
if total == 0:
cv2.putText(frame,"Fake",(int((rect[0]+rect[2])/2),rect[1]+rect[3]),cv2.FONT_HERSHEY_SIMPLEX,1,(0,0,255),2)
else:
cv2.putText(frame,"Blink:{}".format(total),(int((rect[0]+rect[2])/2),rect[1]+rect[3]),cv2.FONT_HERSHEY_SIMPLEX,1,(0,0,255),2)
if detected_face[i][0].lower() == 'akash' and count==0 and total >=1:
count+=1
daily_attendance.loc[daily_attendance['Name']=='Akash Yadav','Attendance']=count
elif detected_face[i][0].lower() == 'pawan' and count==0:
count+=1
daily_attendance.loc[daily_attendance['Name']=='Pawan Jha','Attendance']=count
cv2.imshow("Face_Recognition",frame)
ch = cv2.waitKey(1)
#end_time = time.time()
#(end_time - start_time)/60 >= 2 or
if ch == ord('q'):
cap.release()
cv2.destroyAllWindows()
print("Face recognition completed.")
break
attendance['Attendance']+=daily_attendance['Attendance']
attendance.to_csv(r'.\Attendance.csv',index=False)
def e_a_r(eye):
#vertical distance
A = dist.euclidean(eye[1],eye[5])
B = dist.euclidean(eye[2],eye[4])
#horizontal distance
C = dist.euclidean(eye[0],eye[3])
#eye aspect ratio
ear = (A + B)/(C*2.0)
return ear
def findPeople(feature_attr,positions,thres = 0.6,percent_thres = 70):
'''
feature_attr:- it contain 128D / vector values of face
positions:- it contain position of face on screen it can be either left,right or center
thres:-it is distance threshold
percent_thres = at least 70% feature match with store data
return name/ID and percentage
'''
file = open(r'.\face_feature_dataset_128D.txt','r')
face_dataset = json.loads(file.read())
results = []
for i,feat_128D in enumerate(feature_attr):
result = "Unkonwn"
smallest = sys.maxsize #it hold maxsize of any object or limit that is store by list,str,dict
for person in face_dataset.keys():
person_data = face_dataset[person][positions[i]]#it hold 128D vector data of dataset or stored data in database/disk
for data in person_data:#it will extract list of list of data from list of list
distance = np.sqrt(np.sum(np.square(data-feat_128D)))#it return euclidian distance of original data that is stored in disk -(minus) current data(128D) from camera
if distance < smallest:
smallest = distance #it check for euclidian distance if differece of distance is high than Unkonwn person or if in range of sys.maxsize than it will known person
result = person
percentage = min(100,100*thres/smallest)
if percentage <= percent_thres:
result = "Unkonwn"
results.append((result,percentage))
return results
def new_data_entry():
cap = cv2.VideoCapture(0)
name_id = input("Enter name of ID or yours=> ")
file = open(r'.\face_feature_dataset_128D.txt','r')
dataset = json.loads(file.read())
image_matrix = {'Left':[],'Right':[],'Center':[]}
feature_128D = {'Left':[],'Right':[],'Center':[]}
print("For Training turn your face slowly from left to right and after completion press q to save in file=>")
while True:
ret,frame = cap.read()
rects,landmarks = face_detect.detect_face(frame,80)#80 is min size
for (i,rects) in enumerate(rects):
aligned_face,pos = aligner.align(160,frame,landmarks[i])
if len(aligned_face[0])==160 and len(aligned_face) == 160:
image_matrix[pos].append(aligned_face)
cv2.imshow("Captured Face",cv2.resize(aligned_face,(200,200)))
ch = cv2.waitKey(1)
if ch == ord('q'):
cap.release()
cv2.destroyAllWindows()
break
for pos in image_matrix:
#this line is for multiple channel image :- to convert 128D vector
feature_128D[pos] = [np.mean(extract_feature.get_features(image_matrix[pos]),axis=0).tolist()]
dataset[name_id] = feature_128D
file = open(r'.\face_feature_dataset_128D.txt','w')
file.write(json.dumps(dataset))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument("--mode",type=str,help="Run camera for face recognition",default = 'camera')
args = parser.parse_args(sys.argv[1:])
default_graph = FaceRecGraph()#it take info of data flow in tensorflow's layer
aligner = AlignCustom()
extract_feature = FaceFeature(default_graph)
face_detect = MTCNNDetect(default_graph,scale_factor = 2)#rescale image for fast detection of image
main(args)
#-- the end--#