def _set_face_recognition_model(self):
fr = self.getConfig('dataset','dataset_face_recognition')
if os.path.isfile(fr):
self.face_recognition_model = dlib.face_recognition_model_v1(fr)
else:
print(fr + " is invalid, pls check your config file.")
logging.error(fr + " is invalid, pls check your config file.")
def dlib_generate_500_without_detect(query_paths, dataset_paths, nodetected_dict):
predictor_dat_path = r"./shape_predictor_68_face_landmarks.dat"
# face_rec_dat_path = r"./orig_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./1000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./4000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./10000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./20000_dlib_face_recognition_resnet_model_v1.dat"
face_rec_dat_path = r"./origplus20000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./10x10origplus20000_dlib_face_recognition_resnet_model_v1.dat"
shape_predictor = dlib.shape_predictor(predictor_dat_path)
face_rec = dlib.face_recognition_model_v1(face_rec_dat_path)
embedding_dict = {}
for i, img_path in enumerate(query_paths + dataset_paths):
img = io.imread(img_path)
key = os.path.basename(img_path)
if i % 1000 == 0:
print(i, time.asctime(time.localtime(time.time())))
det = dlib.rectangle(1, 1, 1 + 180, 1 + 180)
shape = shape_predictor(img, det)
face_embedding = face_rec.compute_face_descriptor(img, shape)
embedding_dict[key] = np.array(face_embedding)
print("embedding geranating finished", time.asctime(time.localtime(time.time())))
ground_truth_file = open("ground_truth.txt", "w")
result_file = open("test_result.txt", "w")
qkey_list = []
dkey_list = []
embeddingq = []
embeddingd = []
for qkey in query_paths:
qqkey = os.path.basename(qkey)
for dkey in dataset_paths:
ddkey = os.path.basename(dkey)
qkey_list.append(qqkey)
dkey_list.append(ddkey)
embeddingq.append(embedding_dict[qqkey])
embeddingd.append(embedding_dict[ddkey])
if qqkey[1:] == ddkey[1:]:
ground_truth_file.write(qqkey + "," + ddkey + "\n")
diff = np.subtract(embeddingq, embeddingd)
dist = np.sum(np.square(diff), 1)
for i, _ in enumerate(qkey_list):
result_file.write(qkey_list[i] + "," + dkey_list[i] + "," + str(1.0 / dist[i]) + "\n")
ground_truth_file.close()
result_file.close()
def dlib_generate_embedding(image_paths, actual_issame, embedding_size=128):
predictor_dat_path = r"./shape_predictor_68_face_landmarks.dat"
# face_rec_dat_path = r"./orig_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./1000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./4000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./10000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./20000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./origplus20000_dlib_face_recognition_resnet_model_v1.dat"
face_rec_dat_path = r"./10x10origplus20000_dlib_face_recognition_resnet_model_v1.dat"
# predictor_dat_path = r"C:\sources\dlib\examples\build\Release\shape_predictor_68_face_landmarks.dat"
# face_rec_dat_path = r"C:\sources\dlib\examples\build\Release\orig_dlib_face_recognition_resnet_model_v1.dat"
detector = dlib.get_frontal_face_detector()
shape_predictor = dlib.shape_predictor(predictor_dat_path)
face_rec = dlib.face_recognition_model_v1(face_rec_dat_path)
embedding_list = []
non_only_one_list = []
for i, img_path in enumerate(image_paths):
img = io.imread(img_path)
dets = detector(img, 1)
if i % 1000 == 0:
print(i, time.asctime(time.localtime(time.time())))
if len(dets) != 1:
# print('%s failed on face detection, please check it.' % img_path)
# If we can't detect only one face, we temporarily don't need it.
non_only_one_list.append(i)
embedding_list.append(None)
continue
shape = shape_predictor(img, dets[0])
face_embedding = face_rec.compute_face_descriptor(img, shape)
embedding_list.append(np.array(face_embedding))
non_only_one_list.reverse()
print("skip %i pairs" % len(non_only_one_list))
for _, del_id in enumerate(non_only_one_list):
# !!! be careful here. !!!
# 1) we are using python 2 here, python 3 will NOT handle the following as we need.
# 2) the order is important on manipulating the list.
real_id = del_id / 2
del actual_issame[real_id]
del embedding_list[real_id*2 + 1]
del embedding_list[real_id*2]
return embedding_list, actual_issame
def __init__(self, landmarks=None, embedding=None):
"""Face detection
Parameters
----------
landmarks : str
Path to dlib's 68 facial landmarks predictor model.
embedding : str
Path to dlib's face embedding model.
"""
super(Face, self).__init__()
# face detection
self.face_detector_ = dlib.get_frontal_face_detector()
# landmark detection
if landmarks is not None:
self.shape_predictor_ = dlib.shape_predictor(landmarks)
# face embedding
if embedding is not None:
self.face_recognition_ = dlib.face_recognition_model_v1(embedding)
def __init__(self):
self.cap = cv2.VideoCapture(0)
self.detector = dlib.get_frontal_face_detector()
#self.cascade = cv2.CascadeClassifier("data/haarcascade_frontalface_default.xml")
self.predictor = dlib.shape_predictor("data/shape_predictor_68_face_landmarks.dat")
self.face_rec = dlib.face_recognition_model_v1("data/dlib_face_recognition_resnet_model_v1.dat")
def dlib_para_init(self):
# 人脸关键点检测器
# face_landmark_path = '/home/dx/Desktop/detect_face_pyqt5/shape_predictor_68_face_landmarks.dat'
face_landmark_path = os.path.join(os.path.dirname(os.path.abspath('__file__')),
'shape_predictor_68_face_landmarks.dat')
# 人脸识别模型
# face_recognize_path = '/home/dx/Desktop/detect_face_pyqt5/dlib_face_recognition_resnet_model_v1.dat'
face_recognize_path = os.path.join(os.path.dirname(os.path.abspath('__file__')),
'dlib_face_recognition_resnet_model_v1.dat')
# 候选人文件夹
# faces_folder_path = '/home/dx/Desktop/detect_face_pyqt5/candidate-faces'
self.faces_folder_path = os.path.join(os.path.dirname(os.path.abspath('__file__')),
'candidate-faces/')
# 1.加载正脸检测器
self.detector = dlib.get_frontal_face_detector()
# 2.加载人脸关键点检测器
self.landmark = dlib.shape_predictor(face_landmark_path)
# 3. 加载人脸识别模型
self.facerec = dlib.face_recognition_model_v1(face_recognize_path)
# 候选人脸描述子list
self.descriptors = []
# 对文件夹下的每一个人脸进行:
# 1.人脸检测
# 2.关键点检测
# 3.描述子提取
time_flag = [] # 获取照片中时间
file_glob = os.path.join(self.faces_folder_path, "*.jpg")
file_list = []
file_list.extend(glob.glob(file_glob))
print(file_list)
for i in range(0, len(file_list)):
tmp = str(file_list[i])
tmp_1 = tmp[51:65] # 截取字符串,截取时间
time_flag.append(tmp_1)
#print(time_flag)
cand_d = dict(zip(file_list, time_flag))
cand_sorted = sorted(cand_d.items(), key=lambda d: d[1]) # 按字典的第二个关键字排序
# print(cand_sorted)
for f in range(0, len(cand_sorted)):
print("Processing file: {}".format(cand_sorted[f][0]))
self.img = io.imread(cand_sorted[f][0])
# win.clear_overlay()
# win.set_image(img)
# 1.人脸检测
dets = self.detector(self.img, 1) # 人脸标定
print("Number of faces detected: {}".format(len(dets)))
for k, d in enumerate(dets):
# 2.关键点检测
shape = self.landmark(self.img, d)
# 画出人脸区域和和关键点
# win.clear_overlay()
# win.add_overlay(d)
# win.add_overlay(shape)
# 3.描述子提取，128D向量
face_descriptor = self.facerec.compute_face_descriptor(self.img, shape)
# 转换为numpy array
v = np.array(face_descriptor)
self.descriptors.append(v)
# 对需识别人脸进行同样处理
# 提取描述子，不再注释
# 候选人名单
# self.candidate = ['dwh', 'whr', 'zjr', 'dx', 'dx', 'whr', 'dx']
self.candidate = self.readfile() # 读取候选人,从name.txt中
print(self.candidate)
return np.empty((0))
return np.linalg.norm(face_encodings - face_to_compare, axis=1)
def get_coords(p1):
try:
return int(p1[0][0][0]), int(p1[0][0][1])
except:
return int(p1[0][0]), int(p1[0][1])
face_detector = dlib.get_frontal_face_detector()
predictor_model = "shape_predictor_68_face_landmarks.dat"
face_pose_predictor = dlib.shape_predictor(predictor_model)
face_encoder = dlib.face_recognition_model_v1(
'dlib_face_recognition_resnet_model_v1.dat')
lk_params = dict(winSize=(15, 15),
maxLevel=2,
criteria=(cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, 10,
0.03))
max_head_movement = 20
movement_threshold = 50
gesture_threshold = 175
face_found = False
frame_num = 0
cap = cv2.VideoCapture(0)
while frame_num < 30:
# Take first frame and find corners in it
frame_num += 1
def reco(content):
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor(
'model/shape_predictor_68_face_landmarks.dat')
facerec = dlib.face_recognition_model_v1(
"model/dlib_face_recognition_resnet_model_v1.dat")
now_time = time.time()
while cap.isOpened():
flag, im_rd = cap.read()
dets = detector(im_rd)
tip = '请有序打卡，不要多人一起打卡，看到自己打卡成功即可离开'
im_rd = cv2ImgAddText(im_rd,
tip,
0,
100,
textColor=(0, 0, 255),
textSize=23)
if len(dets) != 0:
biggest_face = dets[0]
# 取占比最大的脸
maxArea = 0
for det in dets:
w = det.right() - det.left()
h = det.top() - det.bottom()
if w * h > maxArea:
biggest_face = det
maxArea = w * h
cv2.rectangle(im_rd,
tuple([biggest_face.left(),
biggest_face.top()]),
tuple([biggest_face.right(),
biggest_face.bottom()]), (255, 0, 0), 2)
# 获取当前捕获到的图像的所有人脸的特征，存储到 features_cap_arr
if (time.time() - now_time) > 2:
shape = predictor(im_rd, biggest_face)
global features_cap
features_cap = facerec.compute_face_descriptor(im_rd, shape)
for key, value in database.items():
compare = return_euclidean_distance(features_cap, value)
if compare == "same": # 找到了相似脸
study_time, ret = take_time(key)
if ret == 1:
content.setText(
str(key + '签到打卡成功现在时间为： ' +
str(datetime.datetime.now())))
elif ret == 2:
content.setText(
str(key + '签退打卡成功现在时间为： ' +
str(datetime.datetime.now())))
elif ret == 3:
content.setText(
str(key + '已经签到打卡成功现在时间为： ' +
str(datetime.datetime.now())) + '请勿重复打卡！')
elif ret == 4:
content.setText(
str(key + '已经签退打卡成功现在时间为： ' +
str(datetime.datetime.now())) + '请勿重复打卡！')
if (not (dic_time.__contains__(key))):
dic_time[key] = [study_time]
else:
if (study_time != 0):
dic_time[key].append(study_time)
now_time = time.time()
cv2.imshow('face_recoinition', im_rd)
if cv2.waitKey(1) & 0xFF == ord('q'):
now_study_time = {}
a = []
b = []
for key, value in dic_time.items():
now_study_time[key] = sec_to_data(sum(value))
print(now_study_time)
for key, values in now_study_time.items():
a.append(key)
b.append(values)
dataframe = pd.DataFrame({'name': a, 'time': b})
now = datetime.datetime.now()
str2 = str(now.month) + '-' + str(now.day)
fd = open((str2 + '.csv'), mode="w")
dataframe.to_csv((str2 + '.csv'), encoding="utf_8_sig")
fd.close()
break
cap.release()
cv2.destroyAllWindows()
from PIL import Image, ImageDraw, ImageFont
from subprocess import call
import firebase_admin
from firebase_admin import credentials
from firebase_admin import firestore
from gpiozero import LED
from time import sleep
# Dlib 정방향 얼굴 인식 detector
detector = dlib.get_frontal_face_detector()
#Dlib 얼굴 landmark 검출기
predictor = dlib.shape_predictor('data/data_dlib/shape_predictor_68_face_landmarks.dat')
#Dlib Resnet 얼굴 인식 모델，128D 특징 벡터 추출
face_reco_model = dlib.face_recognition_model_v1("data/data_dlib/dlib_face_recognition_resnet_model_v1.dat")
#caffe 얼굴 탐지기 로딩
protoPath = os.path.join("face_detector", "deploy.prototxt")
modelPath = os.path.join("face_detector", "res10_300x300_ssd_iter_140000.caffemodel")
net = cv2.dnn.readNetFromCaffe(protoPath, modelPath)
# 진짜 얼굴 탐지 모델 및 레이블 로딩
model = load_model("liveness.model")
le = pickle.loads(open("le.pickle", "rb").read())
# load OpenCV's Haar cascade for face detection (which is faster than
# dlib's built-in HOG detector, but less accurate), then create the facial landmark predictor
print("[INFO] loading facial landmark predictor...")
detector_eye = cv2.CascadeClassifier("haarcascade_frontalface_default.xml")
def main(mode='test', img_path='def'):
t = time.clock()
classes = [
'MXG', 'Sanaken', 'Zofinka', 'Toalk', 'Zissxzirsziiss', 'kiasummer'
]
known_face_encodes = [
np.loadtxt(MAIN_PATH + '/persons/MXG/fv.txt'),
np.loadtxt(MAIN_PATH + '/persons/Sanaken/fv.txt'),
np.loadtxt(MAIN_PATH + '/persons/Zofinka/fv.txt'),
np.loadtxt(MAIN_PATH + '/persons/Toalk/fv.txt'),
np.loadtxt(MAIN_PATH + '/persons/Zissxzirsziiss/fv.txt'),
np.loadtxt(MAIN_PATH + '/persons/kiasummer/fv.txt')
]
known_face_encodes = np.reshape(known_face_encodes, (6, 5, 128))
# get image
if img_path == 'def':
image = cv2.imread('team.jpg', 1)
else:
image = cv2.imread(img_path, 1)
# output
init_align_faces = []
out_arr = []
# get bboxes
fd = FaceDetector()
conf, faceboxes, mboxes = fd.get_faceboxes(image)
color = [255, 255, 255]
# border widths; I set them all to 150
top, bottom, left, right = [120] * 4
image = cv2.copyMakeBorder(image,
top,
bottom,
left,
right,
cv2.BORDER_CONSTANT,
value=color)
bbox_mark_image = image.copy()
# get alignment model
predictor_model = MAIN_PATH + "/models/shape_predictor_68_face_landmarks.dat"
face_pose_predictor = dlib.shape_predictor(predictor_model)
face_aligner = openface.AlignDlib(predictor_model)
# init predection model
predictor_5_point_model = face_recognition_models.pose_predictor_five_point_model_location(
)
pose_predictor_5_point = dlib.shape_predictor(predictor_5_point_model)
face_recognition_model = face_recognition_models.face_recognition_model_location(
)
face_encoder = dlib.face_recognition_model_v1(face_recognition_model)
for i in range(len(faceboxes)):
# get dlib rectangle from facebox
face_rect = dlib.rectangle(faceboxes[i][0], faceboxes[i][1],
faceboxes[i][2], faceboxes[i][3])
# Get the the face's pose
pose_landmarks = face_pose_predictor(image, face_rect)
# Use openface to calculate and perform the face alignment
alignedFace = face_aligner.align(
534,
image,
face_rect,
landmarkIndices=openface.AlignDlib.OUTER_EYES_AND_NOSE)
alignedFace_out = cv2.resize(alignedFace,
(faceboxes[i][2] - faceboxes[i][0],
faceboxes[i][3] - faceboxes[i][1]))
initFace_out = image[faceboxes[i][1]:faceboxes[i][3],
faceboxes[i][0]:faceboxes[i][2]]
init_align_faces.append([initFace_out, alignedFace_out])
# draw marks
parts = dlib.full_object_detection.parts(pose_landmarks)
FaceDetector.draw_marks(bbox_mark_image, parts)
# get face landmarks for feature extraction
landmark_set = pose_predictor_5_point(
alignedFace,
dlib.rectangle(0, 0, alignedFace.shape[0], alignedFace.shape[1]))
# get feature vector
feature_vector = np.array(
face_encoder.compute_face_descriptor(alignedFace, landmark_set, 1))
# known_face_encode = np.loadtxt('persons/MXG/fv.txt')
ind = compare_faces(known_face_encodes, feature_vector)
if (ind != -1):
face_class = classes[ind]
colour = (0, 255, 0)
else:
face_class = "Unknown"
colour = (0, 0, 255)
sh = max(image.shape[0], image.shape[1])
mult = sh / 500
if mult < 1:
mult = 1
cv2.putText(image, face_class, (faceboxes[i][0], faceboxes[i][1] - 10),
cv2.FONT_HERSHEY_SIMPLEX, 0.5 * mult, colour, 1)
cv2.rectangle(bbox_mark_image, (faceboxes[i][0], faceboxes[i][1]),
(faceboxes[i][2], faceboxes[i][3]), (0, 255, 0))
cv2.rectangle(image, (faceboxes[i][0], faceboxes[i][1]),
(faceboxes[i][2], faceboxes[i][3]), (0, 255, 0))
#cv2.rectangle(image, (mboxes[i][0], mboxes[i][1]), (mboxes[i][2], mboxes[i][3]), (0, 255, 0))
#out_arr.append({'x1': faceboxes[i][0], 'y1': faceboxes[i][1], 'x2': faceboxes[i][2], 'y2': faceboxes[i][3], 'class': face_class, 'conf': conf[i]})
out_arr.append({
'x1': mboxes[i][0],
'y1': mboxes[i][1],
'x2': mboxes[i][2],
'y2': mboxes[i][3],
'class': face_class,
'conf': conf[i]
})
t = time.clock() - t
out_imgs = [image, bbox_mark_image, init_align_faces, t]
if mode == 'test':
cv2.imshow("Preview", image)
cv2.waitKey(0)
cv2.destroyAllWindows()
if mode == 'metr':
#cv2.imshow("Preview", image)
#cv2.waitKey(0)
#cv2.destroyAllWindows()
return out_arr
if mode == 'process':
return out_imgs
if mode == 'def':
return out_imgs
import dlib
import cv2
import numpy as np
predictor = dlib.shape_predictor("./saved_model/shape_predictor_68_face_landmarks.dat")
face_rec_model = dlib.face_recognition_model_v1("./saved_model/dlib_face_recognition_resnet_model_v1.dat")
detector = dlib.get_frontal_face_detector()
frame = cv2.imread(r'./images/person.JPG')
face_rect = detector(frame, 1)[0]
shape = predictor(frame, face_rect)
for pt in shape.parts():
pt_pos = (pt.x, pt.y)
cv2.circle(frame, pt_pos, 2, (0, 255, 0), 1)
cv2.imshow("image", frame)
encoding=face_rec_model.compute_face_descriptor(frame, shape)
encoding=np.array(encoding)
print(encoding)
cv2.waitKey()
print(
"Please install `face_recognition_models` with this command before using `face_recognition`:"
)
print()
print(
"pip install git+https://github.com/ageitgey/face_recognition_models")
quit()
face_detector = dlib.get_frontal_face_detector()
predictor_model = face_recognition_models.pose_predictor_model_location()
pose_predictor = dlib.shape_predictor(predictor_model)
face_recognition_model = face_recognition_models.face_recognition_model_location(
)
face_encoder = dlib.face_recognition_model_v1(face_recognition_model)
def _rect_to_css(rect):
"""
Convert a dlib 'rect' object to a plain tuple in (top, right, bottom, left) order
:param rect: a dlib 'rect' object
:return: a plain tuple representation of the rect in (top, right, bottom, left) order
"""
return rect.top(), rect.right(), rect.bottom(), rect.left()
def _css_to_rect(css):
"""
Convert a tuple in (top, right, bottom, left) order to a dlib `rect` object
import dlib
import cv2
import numpy as np
from datetime import datetime
#C:\Users\haneu\fin_pro
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor(
'C:/Users/haneu/fin_pro/modle/shape_predictor_68_face_landmarks.dat')
face_recog = dlib.face_recognition_model_v1(
'C:/Users/haneu/fin_pro/modle/dlib_face_recognition_resnet_model_v1.dat')
#인코딩이 되어있거 가져옴
#print("20개의 인코딩")
#descs = np.load('C:/Users/haneu/fin_pro/a_project/a_descs.npy', allow_pickle=True)[()]
#descs = np.load('C:/Users/haneu/fin_pro/a_project/a5_descs.npy', allow_pickle=True)[()]
#descs = np.load('C:/Users/haneu/fin_pro/a_project/a10_descs.npy', allow_pickle=True)[()]
descs = np.load('C:/Users/haneu/fin_pro/a_project/a20_descs.npy',
allow_pickle=True)[()]
#들어온 영상에서 이미지 찾기
def encode_faces(image):
faces = detector(image, 1)
if len(faces) == 0:
return np.empty(0)
for k, d in enumerate(faces):
land = predictor(image, d)
pic_list.append(img)
except IOError:
print('read error')
return False
else:
print('read successfully')
return pic_name_list, pic_list
# 人脸检测器
detector = dlib.get_frontal_face_detector()
# 关键点检测器
feature_point = dlib.shape_predictor(predictor_path)
# 人脸参数模型
feature_model = dlib.face_recognition_model_v1(face_rc_model_path)
# 候选人特征向量列表
descriptors = []
if __name__ == '__main__':
name_list, pic_list = read_data(face_folder_path)
num = 1
for i in pic_list:
# 人脸检测
dets = detector(i, 1)
for k, d in enumerate(dets):
# 关键点检测
shape = feature_point(i, d)
def get_face_locations(img, times_to_upsample):
locations_rect = face_detector(img, times_to_upsample)
locations_tuple = [
mmodrect_to_tuple(face_location) for face_location in locations_rect
]
return locations_rect, locations_tuple
dirname = os.path.dirname(__file__)
face_detector = dlib.get_frontal_face_detector()
face_detector_cnn = dlib.cnn_face_detection_model_v1(
dirname + '/model/mmod_human_face_detector.dat')
face_landmark_5_points = dlib.shape_predictor(
dirname + '/model/shape_predictor_5_face_landmarks.dat')
face_encoder = dlib.face_recognition_model_v1(
dirname + '/model/dlib_face_recognition_resnet_model_v1.dat')
keep_running = True
guolong_image = cv2.imread(dirname + '/images/guolong.jpg')[:, :, ::-1]
_, guolong_encoding = get_locations_and_encodings(guolong_image)
known_face_encodings = [guolong_encoding[0]]
known_face_names = ['Guolong']
def face_recognition(frame):
small_frame = cv2.resize(frame, (0, 0), fx=0.5, fy=0.5)
rgb_small_frame = small_frame[:, :, ::-1]
w) # Рассчитываем соотношение высоты и построить размеры
dim = (width, int(h * r))
resized = cv2.resize(image, dim,
interpolation=inter) # Меняем размер изображения
return resized # Возвращаем измененный размер изображения
dat_path = 'Data/'
mp3_pic_path = 'mp3_pic/'
detector = dlib.get_frontal_face_detector(
) # создаем объект, который выделяет лицо прямоугольником
predictor = dlib.shape_predictor(
dat_path + 'shape_predictor_68_face_landmarks.dat'
) # загрузка(шаблона) данных обучения для точек на лице
facerec = dlib.face_recognition_model_v1(
dat_path + 'dlib_face_recognition_resnet_model_v1.dat'
) # загрузка данных обучения нейросети resnet
video_capture = cv2.VideoCapture(0) # подключение камеры
video_capture.set(3, 360) # задаем размеры кадра камеры 160x120
video_capture.set(4, 240) # 360x240
img_path = ['p_D.jpg', 'p_N.jpg', 'p_I.jpg', 'p_V.jpg'] #'p_V.jpg'
fdi = []
audio = ['u_D.mp3', 'u_N.mp3', 'u_I.mp3', 'u_V.mp3', 'u_U.mp3']
names = ['Даня', 'Никита', 'Белла', 'Вадимчик']
ind_u = 0
mixer.init()
for im in img_path:
img = cv2.imread(mp3_pic_path + im)
detections = detector(img, 1) # ф-ция выделяет лицо в прямоугольник
for k, d in enumerate(
detections): # цикл по всем найденным на изображении лицам
f = open('/var/www/html/main_pro/upload/data.txt','r')
data = sj.loads(f.read())
f.close()
fo = open('data.txt', 'r')
brr = sj.loads(fo.read())
fo.close()
a = data['name']
b = data['number']
mdl = fm.pose_predictor_model_location()
frm = fm.face_recognition_model_location()
cnn = dlib.face_recognition_model_v1(frm)
fd = dlib.get_frontal_face_detector()
im = cv.imread('test.jpg')
hi = (200,200)
im = cv.resize(im, hi, interpolation = cv.INTER_AREA)
df = fd(im,1)
if len(df)!=0:
mk = 'mkdir '+a
os.system(mk)
fpose = dlib.shape_predictor(mdl)
pl = fpose(im,df[0])
" ./face_recognition.py shape_predictor_68_face_landmarks.dat dlib_face_recognition_resnet_model_v1.dat ../examples/faces\n"
"You can download a trained facial shape predictor and recognition model from:\n"
" http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2\n"
" http://dlib.net/files/dlib_face_recognition_resnet_model_v1.dat.bz2")
exit()
predictor_path = sys.argv[1]
face_rec_model_path = sys.argv[2]
faces_folder_path = sys.argv[3]
# Load all the models we need: a detector to find the faces, a shape predictor
# to find face landmarks so we can precisely localize the face, and finally the
# face recognition model.
detector = dlib.get_frontal_face_detector()
sp = dlib.shape_predictor(predictor_path)
facerec = dlib.face_recognition_model_v1(face_rec_model_path)
win = dlib.image_window()
# Now process all the images
for f in glob.glob(os.path.join(faces_folder_path, "*.jpg")):
print("Processing file: {}".format(f))
img = io.imread(f)
win.clear_overlay()
win.set_image(img)
# Ask the detector to find the bounding boxes of each face. The 1 in the
# second argument indicates that we should upsample the image 1 time. This
# will make everything bigger and allow us to detect more faces.
dets = detector(img, 1)
def dlib_face_cluster(pic_path,tmp_path,files_path):
global detector,sp,facerec
predictor_path = files_path + '/shape_predictor_5_face_landmarks.dat'
face_rec_model_path = files_path + '/dlib_face_recognition_resnet_model_v1.dat'
faces_folder_path = pic_path
output_folder_path = tmp_path
# Load all the models we need: a detector to find the faces, a shape predictor
# to find face landmarks so we can precisely localize the face, and finally the
# face recognition model.
if not 'detector' in globals():
detector = dlib.get_frontal_face_detector()
if not 'sp' in globals():
sp = dlib.shape_predictor(predictor_path)
if not 'facerec' in globals():
facerec = dlib.face_recognition_model_v1(face_rec_model_path)
raw_image_list = glob.glob(os.path.join(faces_folder_path, "*.jpg"))
raw_image_list.append(raw_image_list[0])
counter = 1
batch_index = 1
# Now find all the faces and compute 128D face descriptors for each face.
while(True):
descriptors = []
images = []
while(True):
# print("Processing file: {}".format(f))
try:
f = raw_image_list.pop(0)
counter += 1
except:
break
try:
img = io.imread(f)
except:
print('[Debug] : something wrong at io.imread, {0}'.format(f))
continue
# Ask the detector to find the bounding boxes of each face. The 1 in the
# second argument indicates that we should upsample the image 1 time. This
# will make everything bigger and allow us to detect more faces.
try:
dets = detector(img, 1)
except:
print('[Debug] : something wrong at detector, {0}'.format(f))
continue
# print("Number of faces detected: {}".format(len(dets)))
# Now process each face we found.
for k, d in enumerate(dets):
# Get the landmarks/parts for the face in box d.
shape = sp(img, d)
# Compute the 128D vector that describes the face in img identified by
# shape.
try:
face_descriptor = facerec.compute_face_descriptor(img, shape)
descriptors.append(face_descriptor)
images.append((img, shape))
except:
print('[Debug] : something wrong at facerec.compute_face_descriptor, {0}'.format(f))
continue
# Now let's cluster the faces.
labels = dlib.chinese_whispers_clustering(descriptors, 0.5)
num_classes = len(set(labels))
# print("Number of clusters: {}".format(num_classes))
# Find biggest class
biggest_class = None
biggest_class_length = 0
for i in range(0, num_classes):
class_length = len([label for label in labels if label == i])
if class_length > biggest_class_length:
biggest_class_length = class_length
biggest_class = i
# print("Biggest cluster id number: {}".format(biggest_class))
# print("Number of faces in biggest cluster: {}".format(biggest_class_length))
# Find the indices for the biggest class
indices = []
for i, label in enumerate(labels):
if label == biggest_class:
indices.append(i)
# print("Indices of images in the biggest cluster: {}".format(str(indices)))
# Ensure output directory exists
if not os.path.isdir(output_folder_path):
os.makedirs(output_folder_path)
# Save the extracted faces
# print("Saving faces in largest cluster to output folder...")
for i, index in enumerate(indices):
img, shape = images[index]
file_path = os.path.join(output_folder_path,'batch{0}_{1}'.format(batch_index,i))
dlib.save_face_chip(img, shape, file_path)
break
batch_index += 1
if len(raw_image_list) == 0:
break
def process_video(folder, file_name=None):
video_src = os.path.join(os.curdir, folder, file_name)
images_folder = os.path.join(folder, 'frames')
if not os.path.isdir(images_folder):
os.mkdir(images_folder)
if os.path.isfile(video_src):
print('it is a file: ', video_src)
else:
print('no file you n00b', video_src)
predictor_path = "shape_predictor_68_face_landmarks.dat"
face_rec_model_path = "dlib_face_recognition_resnet_model_v1.dat"
detector = dlib.get_frontal_face_detector()
sp = dlib.shape_predictor(predictor_path)
facerec = dlib.face_recognition_model_v1(face_rec_model_path)
cam = cv2.VideoCapture(video_src)
image_count = 0
frame_count = 0
while True:
frame_count += 1
if frame_count % 1000 == 1:
print(
'frames processed: ', frame_count - 1
) # yes, it's off by one, but looks better with whole thousands
ret, img = cam.read()
if not ret:
break
if not int(cam.get(cv2.CAP_PROP_POS_FRAMES)) % 15 == 0:
continue
# Ask the detector to find the bounding boxes of each face. The 1 in the
# second argument indicates that we should upsample the image 1 time. This
# will make everything bigger and allow us to detect more faces.
dets = detector(img, 1)
# print("Number of faces detected: {}".format(len(dets)))
if len(dets) > 0:
# Now process each face we found.
for k, d in enumerate(dets):
# print("Detection {}: Left: {} Top: {} Right: {} Bottom: {}".format(
# k, d.left(), d.top(), d.right(), d.bottom()))
# Get the landmarks/parts for the face in box d.
shape = sp(img, d)
# Compute the 128D vector that describes the face in img identified by
# shape.
face_id = facerec.compute_face_descriptor(img, shape)
if not have_face(face_id):
print("new face found at frame number {0}".format(
frame_count))
face_list.append(
facerec.compute_face_descriptor(img, shape))
save_file_to = os.path.join(
images_folder,
"{0}{1}.jpg".format(file_name[:-4], image_count))
# save image as file
cv2.imwrite(save_file_to, img)
image_count += 1
print("{0} images made out of {1} frames".format(image_count, frame_count))
cv2.destroyAllWindows()
import sys
import dlib
import numpy as np
import os
import glob
from skimage import io
import re
current_path = os.getcwd() #获取当前路径
predictor_path = current_path + '/model/shape_predictor_68_face_landmarks.dat' # 1.人脸关键点检测器
face_rec_model_path = current_path + '/model/dlib_face_recognition_resnet_model_v1.dat' # 2.人脸识别模型
face_folder_path = current_path + '/all_face' # 3.所有人脸文件夹
detector = dlib.get_frontal_face_detector() #加载正脸识别器
shaper = dlib.shape_predictor(predictor_path) #加载人脸关键识别器
facerec = dlib.face_recognition_model_v1(face_rec_model_path) #加载人脸模型
all_face_d = {}
for img_path in glob.glob(os.path.join(face_folder_path, "*.jpg")):
print("processing file: {}".format(img_path))
a = os.path.basename(img_path)
b = re.compile('\d+')
filename = b.findall(a)[0]
print(filename)
img = io.imread(img_path)
dets = detector(img, 1)
print("number of faces detected: {}".format(len(dets)))
for index, face in enumerate(dets):
shape = shaper(img, face) # 2.关键点检测
face_desciptor = facerec.compute_face_descriptor(
def __init__(self,
image_topic='/usb_cam/image_raw',
predictor_model=pkg_path +
'/models/shape_predictor_68_face_landmarks.dat',
recognition_model=pkg_path +
'/models/dlib_face_recognition_resnet_model_v1.dat',
database_names=os.listdir(pkg_path + '/training_data'),
n_input=128,
n_output=len(os.listdir(pkg_path + '/training_data')),
load_model=True,
prob_threshhold=0.97):
# define global variables
self.n_input = n_input
self.n_output = n_output
self.database_names = np.array(database_names)
self.prob_threshold = prob_threshhold
# define network for face prediction
def model(embedding):
l_in = InputLayer(input_var=embedding, shape=(None, n_input))
h_1 = DenseLayer(l_in, num_units=64, nonlinearity=rectify)
h_2 = DenseLayer(h_1, num_units=32, nonlinearity=rectify)
probs = DenseLayer(h_1, num_units=n_output, nonlinearity=softmax)
return probs
self.X = T.fmatrix()
self.Y = T.fmatrix()
self.network_probs = model(self.X)
# load the trained weights for the network
try:
if load_model:
with np.load(pkg_path + '/models/model.npz') as f:
param_values = [
f['arr_%d' % i] for i in range(len(f.files))
]
set_all_param_values(self.network_probs, param_values)
except IOError:
print 'run the training script before use!'
# define the network functions
self.output = get_output(self.network_probs)
self.predict_faces = theano.function(inputs=[self.X],
outputs=self.output,
allow_input_downcast=True)
# define all face recognition functions
self.face_detector = dlib.get_frontal_face_detector()
self.face_pose_predictor = dlib.shape_predictor(predictor_model)
self.face_recognition_model = dlib.face_recognition_model_v1(
recognition_model)
# start the ros pipeline
self.face_features_pub = rospy.Publisher(
'face_recognition/marked_faces', Image, queue_size=0)
self.face_labels_pub = rospy.Publisher(
'face_recognition/labelled_faces', Image, queue_size=0)
self.face_msgs_pub = rospy.Publisher('face_recognition/detected_faces',
DetectedFaces,
queue_size=0)
self.bridge = CvBridge()
self.camera_sub = rospy.Subscriber(image_topic, Image, self.camera_cb)
import argparse
from face_detector import detect_faces
from face_embeddings import extract_face_embeddings
import pickle
import dlib
from tqdm import tqdm
args = parse_args()
embeddings = np.load(args.embeddings)
labels = pickle.load(open(args.labels, 'rb'))
shape_predictor = dlib.shape_predictor(
"models/"
"shape_predictor_5_face_landmarks.dat")
face_recognizer = dlib.face_recognition_model_v1(
"models/"
"dlib_face_recognition_resnet_model_v1.dat")
image_path_list = [l.strip() for l in open(args.input_list).readlines()]
output_dict = dict()
for idx, img_path in enumerate(tqdm(image_path_list)):
try:
image = cv2.imread(img_path)
if image is None:
output_dict[img_path] = []
continue
image_original = image.copy()
image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
faces = detect_faces(image)
def test():
# 取得dlib預設的臉部偵測器
detector = dlib.get_frontal_face_detector()
# 根據shape_predictor方法載入68個特徵點模型，此方法為人臉表情識別的偵測器
predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")
# 臉部校正
fa = FaceAligner(predictor, desiredFaceWidth=256)
# 載入人臉辨識檢測器
facerec = dlib.face_recognition_model_v1(
"dlib_face_recognition_resnet_model_v1.dat")
# 比對人臉描述子列表
descriptors = []
# 比對人臉名稱列表
candidate = []
# 開啟影片檔案
cap = cv2.VideoCapture(0, cv2.CAP_DSHOW)
# 比對人臉圖片資料夾名稱
faces_folder_path = "./rec"
WHO = None
Erro = 0
# 有讀過的照片存成新的檔可直接使用，有新的圖片再加
# 有讀過的照片存成新的檔可直接使用，有新的圖片再加
# 路徑中的file個數
NUM_OF_FILES = 0 # 檔案數
for fn in os.listdir(faces_folder_path):
NUM_OF_FILES += 1
# print(NUM_OF_FILES)
person = [0 for i in range(NUM_OF_FILES)]
# print(person)
# 原本的個數
f = open("numOfFiles.txt", encoding="utf-8")
num_o = f.read()
# print(type(num_o))
num_o = int(num_o)
# print(type(num_o))
# print(num_o)
if num_o != NUM_OF_FILES:
# 把新的個數寫到txt檔裡
num = open('numOfFiles.txt', 'w', encoding="utf-8")
num.write(str(NUM_OF_FILES))
num.close()
# 跑新的
# 讀取資料夾裡的圖片及檔案名(人名)，並將每張圖的128維特徵向量存到description一維陣列中
for f in glob.glob(os.path.join(faces_folder_path, "*.jpg")):
base = os.path.basename(f)
# 依序取得圖片檔案人名，存到candidate一維陣列中
candidate.append(os.path.splitext(base)[0])
img = io.imread(f)
gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
# 1.人臉偵測
face_rects = detector(img, 0)
# print(face_rects)
for index, face in enumerate(face_rects):
# 2.人臉校正
faceAligned = fa.align(img, gray, face)
# 3.再人臉偵測(去除校正後多餘的部分)
face_rects2 = detector(faceAligned, 1)
for index2, face2 in enumerate(face_rects2):
ax1 = face2.left()
ay1 = face2.top()
ax2 = face2.right()
ay2 = face2.bottom()
# 4.68特徵點偵測
shape = predictor(faceAligned, face2)
# 5.取得描述子，128維特徵向量
face_descriptor = facerec.compute_face_descriptor(
faceAligned, shape)
# 轉換numpy array格式
v = np.array(face_descriptor)
descriptors.append(v)
# print(face_rects)
# 存新的到原本的txt檔
np.savetxt('desc_txt.txt', descriptors)
c = open("cand.txt", "w", encoding="utf-8")
str_list = [line + '\n' for line in candidate]
c.writelines(str_list)
c.close()
# 以迴圈從影片檔案讀取影格，並顯示出來
while cap.isOpened():
desc_list = np.loadtxt("./desc_txt.txt")
k = open("cand.txt", encoding="utf-8")
can = k.readlines()
x1 = 0
x2 = 0
y1 = 0
y2 = 0
# 從視訊鏡頭擷取畫面
ret, frame = cap.read()
# 縮小圖片
frame = imutils.resize(frame, width=800)
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
gray = cv2.cvtColor(frame, cv2.COLOR_RGB2GRAY)
# 1.人臉偵測
face_rects = detector(frame, 1)
# 取出所有偵測的結果(所有人臉座標點)
for index, rect in enumerate(face_rects):
x1 = rect.left()
y1 = rect.top()
x2 = rect.right()
y2 = rect.bottom()
# 以方框標示偵測的人臉
cv2.rectangle(frame, (x1, y1), (x2, y2), (0, 255, 0), 4,
cv2.LINE_AA)
# 2.人臉校正
faceAligned = fa.align(frame, gray, rect)
# 3.再人臉偵測(去除校正後多餘的部分)
face_rects2 = detector(faceAligned, 1)
# 取出所有偵測的結果(所有人臉座標點)
for index2, rect2 in enumerate(face_rects2):
ax1 = rect2.left()
ay1 = rect2.top()
ax2 = rect2.right()
ay2 = rect2.bottom()
# 4.68特徵點偵測
shape = predictor(faceAligned, rect2)
# 5.取得描述子，128維特徵向量
face_descriptor = facerec.compute_face_descriptor(
faceAligned, shape)
# 轉換numpy array格式
d_test = np.array(face_descriptor)
# 計算歐式距離 (與圖片庫裡，各個人臉間的距離)(5張照片就有5個距離)
# 清空 存放人臉距離的陣列
dist = []
for index in desc_list:
# 計算距離
dist_ = np.linalg.norm(index - d_test)
# 加入陣列
dist.append(dist_)
# print(dist)
# 辨識人名
if dist != []:
# 將比對人名和比對出來的歐式距離組成一個dict
c_d = dict(zip(can, dist))
# 根據歐式距離由小到大排序 [("名字",距離)]二微陣列
cd_sorted = sorted(c_d.items(), key=lambda d: d[1])
# print(cd_sorted)
# 歐式距離(0~1)越小越像，設定0.5作為最低辨識標準
if cd_sorted[0][1] < 0.4:
rec_name = cd_sorted[0][0]
else:
rec_name = "No Data"
Erro = Erro + 1
if Erro == 6:
WHO = "No Data"
cap.release()
cv2.destroyAllWindows()
break
imgPil = Image.fromarray(frame)
font = ImageFont.truetype("C:/Windows/Fonts/msjh.ttc", 20)
draw = ImageDraw.Draw(imgPil)
draw.fontmode = '1' # 關閉反鋸齒
draw.text((x1, y1 - 20),
rec_name,
font=font,
fill=(255, 255, 255))
frame = np.array(imgPil)
for index, name in enumerate(can):
if rec_name == name:
person[index] = person[index] + 1
# print(person)
if person[index] == 4:
WHO = can[index]
# 標示辨識的人名(中文)
if WHO is not None:
# imgPil.show()
print(WHO)
time.sleep(5)
cap.release()
cv2.destroyAllWindows()
return WHO
break
elif sum(person) > 8:
WHO = "No Data"
cap.release()
cv2.destroyAllWindows()
break
# 標示辨識的人名(只能標示英文)
# cv2.putText(frame, rec_name, (x1, y1), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2, cv2.LINE_AA)
# CV2是用BGR
frame = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)
"""
# FPS
# Find OpenCV version
(major_ver, minor_ver, subminor_ver) = (cv2.__version__).split('.')
if int(major_ver) < 3:
fps = cap.get(cv2.cv.CV_CAP_PROP_FPS)
print("Frames per second using video.get(cv2.cv.CV_CAP_PROP_FPS): {0}".format(fps))
else:
fps = cap.get(cv2.CAP_PROP_FPS)
print("Frames per second using video.get(cv2.CAP_PROP_FPS) : {0}".format(fps))
"""
# 顯示結果
cv2.imshow("Face Detection", frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cap.release()
cv2.destroyAllWindows()
import zlib
import numpy
import pandas
import PIL.Image
import tkinter as tk
import face_recognition
from PIL import ImageDraw, ImageFont, ImageTk
from PyFaceDet import facedetectcnn
from golbalvar import *
from io import BytesIO
from tkinter import *
from sql import *
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor(shape_predictor_68_face_landmarks_path)
facerec = dlib.face_recognition_model_v1(
dlib_face_recognition_resnet_model_v1_path)
font = ImageFont.truetype(
os.path.abspath(os.path.join(os.getcwd(), "../lib")) + "\simsun.ttc",
30,
encoding='utf-8')
#从数据库中提取的已知的数据
knew_id = []
knew_name = []
knew_face_feature = []
def draw_person_name(num, img_rd, faces, face_position_list, face_name):
"""
绘制名字
:param faces:检测到的人脸
:param img_rd:帧图像
import dlib
import face_recognition
from cv2 import cv2
import os
import numpy as np
shape_predictor_model = "shape_predictor_68_face_landmarks.dat"
face_rec_128 = "dlib_face_recognition_resnet_model_v1.dat"
face_Unique_feature = dlib.face_recognition_model_v1(face_rec_128)
face_detector = dlib.get_frontal_face_detector()
face_pose_predictor = dlib.shape_predictor(shape_predictor_model)
def load_images_from_folder(folder):
features = []
for filename in os.listdir(folder):
img = cv2.imread(os.path.join(folder,filename))
if img is not None:
face_2 = face_detector(img, 1)
pose_landmarks_1 = face_pose_predictor(img, face_2[0])
features.append(face_Unique_feature.compute_face_descriptor(img, pose_landmarks_1))
return features
data = np.array(load_images_from_folder("C:\\Users\\Abdul Jalil\\face"))
np.save('data.npy', data)
def dlib_generate_result(query_paths, dataset_paths, nodetected_dict):
predictor_dat_path = r"./shape_predictor_68_face_landmarks.dat"
# face_rec_dat_path = r"./orig_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./1000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./4000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./10000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./20000_dlib_face_recognition_resnet_model_v1.dat"
face_rec_dat_path = r"./origplus20000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./10x10origplus20000_dlib_face_recognition_resnet_model_v1.dat"
detector = dlib.get_frontal_face_detector()
shape_predictor = dlib.shape_predictor(predictor_dat_path)
face_rec = dlib.face_recognition_model_v1(face_rec_dat_path)
embedding_dict = {}
nodetected_fail = False
for i, img_path in enumerate(query_paths + dataset_paths):
img = io.imread(img_path)
key = os.path.basename(img_path)
try:
dets = detector(img, 1)
except RuntimeError:
print('%s failed on image format, please check it.' % img_path)
continue
if i % 1000 == 0:
print(i, time.asctime(time.localtime(time.time())))
det = None
if len(dets) < 1:
# If we can't detect only one face, we give the position of face.
try:
det = nodetected_dict[key]
except KeyError:
print('%s failed on face detection, please check it.' % img_path)
nodetected_fail = True
if len(dets) >= 1:
det = dets[0]
# print('%s multi-faces detected during face detection, please check it.' % img_path)
# If we detected multi-face, we choose the biggest rectangle.
for k, d in enumerate(dets):
if (d.right() - d.left()) * (d.bottom() - d.top()) > (det.right() - det.left()) * (
det.bottom() - det.top()):
det = d
shape = shape_predictor(img, det)
face_embedding = face_rec.compute_face_descriptor(img, shape)
embedding_dict[key] = np.array(face_embedding)
if nodetected_fail:
sys.exit(1)
print("embedding geranating finished", time.asctime(time.localtime(time.time())))
ground_truth_file = open("j_ground_truth.txt", "w")
result_file = open("j_test_result.txt", "w")
qkey_list = []
dkey_list = []
embeddingq = []
embeddingd = []
for qkey in query_paths:
qqkey = os.path.basename(qkey)
for dkey in dataset_paths:
ddkey = os.path.basename(dkey)
qkey_list.append(qqkey)
dkey_list.append(ddkey)
embeddingq.append(embedding_dict[qqkey])
embeddingd.append(embedding_dict[ddkey])
if qqkey[2:] == ddkey[2:]:
ground_truth_file.write(qqkey + "," + ddkey + "\n")
diff = np.subtract(embeddingq, embeddingd)
dist = np.sum(np.square(diff), 1)
for i, _ in enumerate(qkey_list):
result_file.write(qkey_list[i] + "," + dkey_list[i] + "," + str(1.0 / dist[i]) + "\n")
ground_truth_file.close()
result_file.close()
parser.add_argument("--save_faces")
parser.add_argument("--similarity_threshold", type=float, default=0.5)
parser.add_argument("--video_max_images", type=int, default=10)
parser.add_argument("--type",
choices=['default', 'photobooth', 'google', 'crime'],
default='default')
args = parser.parse_args()
if args.detector == 'hog':
detector = dlib.get_frontal_face_detector()
elif args.detector == 'cnn':
detector = dlib.cnn_face_detection_model_v1(args.cnn_model_path)
else:
assert False, "Unknown detector " + args.detector
predictor = dlib.shape_predictor(args.predictor_path)
facerec = dlib.face_recognition_model_v1(args.face_rec_model_path)
if args.save_resized:
makedirs(args.save_resized)
if args.save_faces:
makedirs(args.save_faces)
db.connect(args.db)
print("Processing files...")
start_time = time.time()
num_files = 0
num_images = 0
num_faces = 0
for dirpath, dirnames, filenames in os.walk(args.dir):
import os
import glob
import cv2
import dlib
import numpy as np
from PIL import Image
detectorFace = dlib.get_frontal_face_detector()
detectorPontos = dlib.shape_predictor(
"recursos/shape_predictor_68_face_landmarks.dat")
reconhecimentoFacial = dlib.face_recognition_model_v1(
"recursos/dlib_face_recognition_resnet_model_v1.dat")
indices = np.load("recursos/indices_yale.pickle")
descritoresFaciais = np.load("recursos/descritores_yale.npy")
limiar = 0.5
totalFaces = 0
totalAcertos = 0
for arquivo in glob.glob(os.path.join("yalefaces/teste", "*.gif")):
imagemFace = Image.open(arquivo).convert('RGB')
imagem = np.asarray(imagemFace, 'uint8')
idatual = int(
os.path.split(arquivo)[1].split(".")[0].replace("subject", ""))
totalFaces += 1
facesDetectadas = detectorFace(imagem, 3)
for face in facesDetectadas:
e, t, d, b = (int(face.left()), int(face.top()), int(face.right()),
int(face.bottom()))
pontosFaciais = detectorPontos(imagem, face)
descritorFacial = reconhecimentoFacial.compute_face_descriptor(
imagem, pontosFaciais)
import dlib, cv2
import numpy as np
detector = dlib.get_frontal_face_detector()
sp = dlib.shape_predictor('models/shape_predictor_68_face_landmarks.dat')
facerec = dlib.face_recognition_model_v1(
'models/dlib_face_recognition_resnet_model_v1.dat')
descs = np.load('img/descs.npy')[()]
def encode_face(img):
dets = detector(img, 1)
if len(dets) == 0:
return np.empty(0)
for k, d in enumerate(dets):
shape = sp(img, d)
face_descriptor = facerec.compute_face_descriptor(img, shape)
return np.array(face_descriptor)
video_path = 'img/bs_clip2.mp4'
cap = cv2.VideoCapture(video_path)
if not cap.isOpened():
exit()
_, img_bgr = cap.read() # (800, 1920, 3)
import numpy as np
import dlib
from imgaug import augmenters as iaa
import pandas as pd
from sklearn.svm import SVC
import random
face_encoder = dlib.face_recognition_model_v1('./model/dlib_face_recognition_resnet_model_v1.dat')
face_pose_predictor = dlib.shape_predictor('./model/shape_predictor_68_face_landmarks.dat')
def encoding_faces(images, label, coord_detect):
"""
Encoding a list of faces using FaceNet generating a 128D vector
:param images: list of images of faces to encode
:param coord_detect: coordinates of the detection
:return: pandas dataframe of the faces for one label
"""
l = []
for img, d in zip(images, coord_detect):
(x1, y1, x2, y2) = d
detected_face = dlib.rectangle(left=0, top=0, right=int(x2-x1), bottom=int(y2-y1))
pose_landmarks = face_pose_predictor(img, detected_face)
face_encoding = face_encoder.compute_face_descriptor(img, pose_landmarks, 1)
l.append(np.append(face_encoding, [label]))
return np.array(l)
def create_positive_set(pictures, coords, label=1):
"""
Create positive train set for one face from a list of three pictures of this face.
Data Augmentation on these three pictures to generate 10 pictures.
import numpy as np
from sklearn import svm
import pickle
import glob
import os
import dlib
import cv2
##Final Variables
predictor_url="/home/shubham/Documents/project/Final_project/data_functioning/shape_predictor_68_face_landmarks.dat"
recog_model_url="/home/shubham/Documents/project/Final_project/data_functioning/dlib_face_recognition_resnet_model_v1.dat"
to_recog="/home/shubham/Documents/project/Final_project/data/"
detector=dlib.get_frontal_face_detector()
sp=dlib.shape_predictor(predictor_url)
facerec=dlib.face_recognition_model_v1(recog_model_url)
files=[]
def get_arr(frame):
dets=detector(frame,1)
if len(dets)!=0:
for l,d in enumerate(dets):
shape=sp(frame,d)
face_descriptor=facerec.compute_face_descriptor(frame,shape)
return list(face_descriptor)
def arr_to_str(arr):
retStr=str(arr[0])
for i in range(1,len(arr)):
retStr=retStr+","+str(arr[i])
return retStr
for f in glob.glob(os.path.join("training_vdos/", "*.mp4")):
files.append(f)
for fl_url in files:
label_processing=fl_url.split("/")[1].split(".")[0]
import os
import dlib
import numpy as np
from skimage import io
import cv2
from collections import namedtuple
data_dir = os.path.expanduser(
'/home/rezeck/Documents/itelicam/face-recognition-dlib/data')
faces_folder_path = data_dir + '/kodemaker/'
# Globals
dlib_frontal_face_detector = dlib.get_frontal_face_detector()
shape_predictor = dlib.shape_predictor(
data_dir + '/dlib/shape_predictor_5_face_landmarks.dat')
face_recognition_model = dlib.face_recognition_model_v1(
data_dir + '/dlib/dlib_face_recognition_resnet_model_v1.dat')
face_classifier_opencv = cv2.CascadeClassifier(
data_dir + '/dlib/haarcascade_frontalface_default.xml')
def to_dlib_rect(w, h):
return dlib.rectangle(left=0, top=0, right=w, bottom=h)
def to_rect(dr):
# (x, y, w, h)
return dr.left(), dr.top(), dr.right() - dr.left(), dr.bottom() - dr.top()
def face_detector_opencv(image):
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
def dlib_generate_embedding(query_paths, dataset_paths):
predictor_dat_path = r"./shape_predictor_68_face_landmarks.dat"
face_rec_dat_path = r"./orig_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./1000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./4000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./10000_dlib_face_recognition_resnet_model_v1.dat"
# face_rec_dat_path = r"./20000_dlib_face_recognition_resnet_model_v1.dat"
# predictor_dat_path = r"C:\sources\dlib\build\Release\shape_predictor_68_face_landmarks.dat"
# face_rec_dat_path = r"C:\sources\dlib\build\Release\orig_dlib_face_recognition_resnet_model_v1.dat"
detector = dlib.get_frontal_face_detector()
shape_predictor = dlib.shape_predictor(predictor_dat_path)
face_rec = dlib.face_recognition_model_v1(face_rec_dat_path)
embedding_dict = {}
non_only_one_list = []
more_than_one_list = []
for i, img_path in enumerate(query_paths + dataset_paths):
img = io.imread(img_path)
key = os.path.basename(img_path)
dets = detector(img, 1)
# win.clear_overlay()
# win.set_image(img)
if i % 1000 == 0:
print(i, time.asctime(time.localtime(time.time())))
det = None
if len(dets) < 1:
# If we can't detect only one face, we give the position of face.
try:
det = nodetected_dict[key]
except KeyError:
print('%s failed on face detection, please check it.' % img_path)
continue
if len(dets) >= 1:
det = dets[0]
# print('%s multi-faces detected during face detection, please check it.' % img_path)
# If we detected multi-face, we choose the biggest rectangle.
for k, d in enumerate(dets):
if (d.right() - d.left()) * (d.bottom() - d.top()) > (det.right() - det.left()) * (det.bottom() - det.top()):
det = d
shape = shape_predictor(img, det)
face_embedding = face_rec.compute_face_descriptor(img, shape)
embedding_dict[key] = np.array(face_embedding)
print("embedding geranating finished", time.asctime(time.localtime(time.time())))
actual_issame = []
embedding_list = []
for qkey in query_paths:
qqkey = os.path.basename(qkey)
for dkey in dataset_paths:
ddkey = os.path.basename(dkey)
embedding_list.append(embedding_dict[qqkey])
embedding_list.append(embedding_dict[ddkey])
if qqkey[1:] == ddkey[1:]:
actual_issame = True
else:
actual_issame = False
return embedding_list, actual_issame
import os
import shutil
import dlib
cluster_size_threshold = 2 # 修改此阈值参数可更改类内最低的人脸数
detector = dlib.get_frontal_face_detector()
sp = dlib.shape_predictor('users/config/shape_predictor_68_face_landmarks.dat')
facerec = dlib.face_recognition_model_v1(
'users/config/dlib_face_recognition_resnet_model_v1.dat')
# 传入所有需要聚类的图片路径
def face_cluster(input_images):
# 初始化，防止多次打开导致重复生成
descriptors = []
images = []
for image in input_images:
# 逐个加载图片
img = dlib.load_rgb_image(image['path'])
# 检测人脸数量和位置
faces = detector(img, 1)
# print(f'{len(faces)} faces detected in', image['path'])
# 遍历所有人脸
for face in faces:
# 利用68关键点模型检测人脸关键点
shape = sp(img, face)
# 将关键点集转化为128维特征向量
face_descriptor = facerec.compute_face_descriptor(img, shape)
help="download at http://dlib.net/files/shape_predictor_68_face_landmarks.dat.bz2")
a.add_argument("--model", default="dlib_face_recognition_resnet_model_v1.dat",
help="download at http://dlib.net/files/dlib_face_recognition_resnet_model_v1.dat.bz2")
a.add_argument("--image", help="images must be .jpg files")
a.add_argument("--image_dir", help="images must be .jpg files")
a.add_argument("--viz_off", action='store_true', help="flag to vizualize detections and shape prediction")
args = a.parse_args()
if args.landmarks_dat is None or args.model is None or (args.image is None and args.image_dir is None):
a.print_help()
sys.exit(1)
detector = dlib.get_frontal_face_detector()
sp = dlib.shape_predictor(args.landmarks_dat)
facerec = dlib.face_recognition_model_v1(args.model)
win = None
if not args.viz_off:
win = dlib.image_window()
if args.image is not None:
img = io.imread(args.image)
print(encode(detector, sp, facerec, img, win))
if args.image_dir is not None:
fns = os.listdir(args.image_dir)
descriptors = []
for fn in fns:
if FILE_EXT not in fn:
continue
import dlib
from skimage import io #这个模块是用来图片输入输出操作的
import csv
path_pics = "D://xx/"
path_csv = "D://xx/"
# detector to find the faces
#使用人脸提取器
detector = dlib.get_frontal_face_detector()
#dlib68点模型
# shape predictor to find the face landmarks
predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat")
# face recognition model, the object maps human faces into 128D vectors#映射到128D的向量中,加载人脸识别模型
facerec = dlib.face_recognition_model_v1(
"dlib_face_recognition_resnet_model_v1.dat")
#返回单张图片的128D特征
def return_128d_features(path_img):
#图片读取
img = io.imread(path_img)
#取灰度
img_gray = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
dets = detector(img_gray, 1)
#检测到人脸
if (len(dets) != 0):
shape = predictor(img_gray, dets[0])
face_descriptor = facerec.compute_face_descriptor(img_gray, shape)
else:
face_descriptor = 0
print("pip install git+https://github.com/ageitgey/face_recognition_models")
quit()
face_detector = dlib.get_frontal_face_detector()
predictor_68_point_model = face_recognition_models.pose_predictor_model_location()
pose_predictor_68_point = dlib.shape_predictor(predictor_68_point_model)
predictor_5_point_model = face_recognition_models.pose_predictor_five_point_model_location()
pose_predictor_5_point = dlib.shape_predictor(predictor_5_point_model)
cnn_face_detection_model = face_recognition_models.cnn_face_detector_model_location()
cnn_face_detector = dlib.cnn_face_detection_model_v1(cnn_face_detection_model)
face_recognition_model = face_recognition_models.face_recognition_model_location()
face_encoder = dlib.face_recognition_model_v1(face_recognition_model)
def _rect_to_css(rect):
"""
Convert a dlib 'rect' object to a plain tuple in (top, right, bottom, left) order
:param rect: a dlib 'rect' object
:return: a plain tuple representation of the rect in (top, right, bottom, left) order
"""
return rect.top(), rect.right(), rect.bottom(), rect.left()
def _css_to_rect(css):
"""
Convert a tuple in (top, right, bottom, left) order to a dlib `rect` object
def main():
if len(sys.argv) != 2:
exit()
os.chdir(sys.path[0])
#欲辨識圖片
img_name = sys.argv[1]
face_data_path = "./face"
# Dlib 的人臉偵測器
detector = dlib.get_frontal_face_detector()
#人臉68特徵點模型檢測器
shape_predictor = dlib.shape_predictor(
"../dat/shape_predictor_68_face_landmarks.dat")
#載入人臉辨識模型及檢測器
face_rec_model = dlib.face_recognition_model_v1(
"../dat/dlib_face_recognition_resnet_model_v1.dat")
descriptors = []
candidate = []
img_files = glob.glob(os.path.join(face_data_path, "*.jpg"))
for file in img_files:
base = os.path.basename(file)
candidate.append(os.path.splitext(base)[0])
img = io.imread(file)
dets = detector(img, 1)
# 取出所有偵測的結果
for k, d in enumerate(dets):
#68特徵點偵測
shape = shape_predictor(img, d)
#128維特徵向量描述
face_descriptor = face_rec_model.compute_face_descriptor(
img, shape)
#轉換numpy array 格式
v = np.array(face_descriptor)
descriptors.append(v)
#辨識圖片處理
img = io.imread(img_name)
face_rects = detector(img, 1)
distance = []
for k, d in enumerate(face_rects):
#68特徵點偵測
shape = shape_predictor(img, d)
#128維特徵向量描述
face_descriptor = face_rec_model.compute_face_descriptor(img, shape)
d_test = np.array(face_descriptor)
x1 = d.left()
y1 = d.top()
x2 = d.right()
y2 = d.bottom()
cv2.rectangle(img, (x1, y1), (x2, y2), (0, 255, 0), 4, cv2.LINE_AA)
for i in descriptors:
#計算向量差值
dist_ = np.linalg.norm(i - d_test)
distance.append(dist_)
candidate_distance_dict = dict(zip(candidate, distance))
#依照距離排序
candidate_distance_dict_sorted = sorted(candidate_distance_dict.items(),
key=lambda d: d[1])
print(candidate_distance_dict_sorted)
#取出最相像的
result = candidate_distance_dict_sorted[0][0]
cv2.putText(img, result, (x1, y1), cv2.FONT_HERSHEY_DUPLEX, 0.7,
(255, 255, 255), 1, cv2.LINE_AA)
#img = imutils.resize(img, width=800)
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
cv2.imshow("test", img)
cv2.waitKey(0)
cv2.destroyAllWindows()
# @Time : 2018/12/4 0004 17:20
# @Author : lzc
# @File : face_recongnition.py
import dlib
import numpy as np
import cv2
import pandas as pd
#首先初始化128d人脸识别模型
recognition = dlib.face_recognition_model_v1('lib/dlib_face_recognition_resnet_model_v1.dat')
#计算向量间的距离
#fea_1是指从特征均值文件中拿到特征数据
#fea_2是指从摄像头拍摄到的人脸特征数据
def get_xiangliang_distance(fea_1,fea_2):
fea_1 = np.array(fea_1)
fea_2 = np.array(fea_2)
distinct = np.sqrt(np.sum(np.square(fea_1-fea_2)))
#根据阈值来判断是否是同一个人，当向量距离差异和大于30%也就是0.3的时候，认为不是同一个人
print('distinct======', distinct)
if distinct > 0.5:
return 'N'
else:
return 'Y'
#首先读取存放入人脸特征的均值
path_face_feature_mean_csv = 'data/feature_mean_all.csv'
csv_rd = pd.read_csv(path_face_feature_mean_csv)
#定义一个集合存放所有人特征均值
def dataGeneration():
print("[INFO] loading facial landmark predictor...")
detector = dlib.get_frontal_face_detector()
predictor = dlib.shape_predictor(pathAttributes.face_detection_model)
face_encoder = dlib.face_recognition_model_v1(
pathAttributes.face_recognition_model)
print("[INFO] camera sensor warming up...")
#vs = VideoStream().start()
#video_capture = cv2.VideoCapture(0)
#time.sleep(2.0)
names = []
s = os.listdir(pathAttributes.faces)
for i in s:
name_w_time = os.path.basename(i)
name_only = name_w_time.split("_", 1)
names.append(name_only)
people = []
IDs = []
for i in s:
name_w_time = os.path.basename(i)
time_only = name_w_time.split("_", 1)[0]
IDs.append(time_only)
#count = 1
for i in s:
for ID in IDs:
if str(i).find(ID) > -1:
document = os.path.join(pathAttributes.faces, i)
#name_w_time = os.path.basename(document)
#name_only = name_w_time.split("_",1)
file_path_data = os.path.join(document, "*.png")
pics = glob.glob(file_path_data)
for index in pics:
pic = os.path.join(document, index)
print(pic)
image = cv2.imread(pic)
try:
faceDetect = detector(image, 1)
shape = predictor(image, faceDetect[0])
face_encoding = face_encoder.compute_face_descriptor(
image, shape, 1)
features = list(face_encoding)
features.insert(0, ID)
#features.insert(1,name_only)
people.append(features)
print(features)
except:
pass
#online mode
newest_data_name = time.strftime("%Y%m%d%H%M%S", time.localtime(
time.time())) + ".csv"
newest_data = os.path.join(pathAttributes.data, newest_data_name)
csvfile = open(newest_data, "w", newline='')
#csvfile = open(pathAttributes.face_features_data,"w",newline='') batch mode
writer = csv.writer(csvfile)
writer.writerows(people)
csvfile.close()
return False
#online mode
"""