def run(self):
# wait for initial detection
while self.shared_variables.detection_box[self.index] is None:
pass
# initiate tracker
self.create_custom_tracker()
# tracking loop
while self.shared_variables.system_running:
self.start_time = datetime.datetime.now()
self.frame = self.shared_variables.frame[self.index]
self.object_custom_tracking()
self.end_time = datetime.datetime.now()
if self.shared_variables.debug:
LOG.debug(
"Tracking time : " + str(self.end_time - self.start_time),
"SYSTEM-" + self.shared_variables.name)
LOG.info("Stopped tracking " + str(self.index),
"SYSTEM-" + self.shared_variables.name)
def run(self):
LOG.info("Start blink frequency "+ str(self.index), "SYSTEM-"+self.shared_variables.name)
# load model
model = load_model('../../model/blinkModel.hdf5')
close_counter = blinks = mem_counter= 0
state = ''
#Wait for detection
while self.shared_variables.frame[self.index] is None:
pass
while self.shared_variables.system_running is not None:
if self.shared_variables.frame[self.index] is not None:
frame = self.shared_variables.frame[self.index]
eyes = self.cropEyes(frame)
if eyes is None:
continue
else:
left_eye,right_eye = eyes
prediction = (model.predict(self.cnnPreprocess(left_eye)) + model.predict(self.cnnPreprocess(right_eye)))/2.0
if prediction > 0.5 :
state = 'open'
close_counter = 0
else:
state = 'close'
close_counter += 1
if state == 'open' and mem_counter > 1:
blinks += 1
mem_counter = close_counter
#save blinking
#eye_state
self.shared_variables.eye_state[self.index] = state
#blinks
self.shared_variables.blinks[self.index] = blinks
#eye_left
self.shared_variables.eye_left[self.index] = left_eye
#eye_right
self.shared_variables.eye_right[self.index] = right_eye
if self.shared_variables.debug:
LOG.debug(str(state) + " " + str(blinks) + " from "+str(self.index),"SYSTEM-"+self.shared_variables.name)
LOG.info("Ending blink freq " + str(self.index), "SYSTEM-"+self.shared_variables.name)
def run(self):
with tf.Session() as sess:
LOG.info("Loading Tensorflow modell " + str(self.index),
"SYSTEM-" + self.shared_variables.name)
# Load model
self.pnet, self.rnet, self.onet = detect_and_align.create_mtcnn(
sess, None)
self.Loaded_model = True
LOG.info("Start tf detections " + str(self.index),
"SYSTEM-" + self.shared_variables.name)
# Start Loop
while self.shared_variables.system_running:
self.start_time = datetime.datetime.now()
frame = self.shared_variables.frame[self.index]
if self.do_flipp_test:
frame = imutils.rotate(
frame, self.flipp_test_degree * self.flipp_test_nr)
# Do detection
face_patches, padded_bounding_boxes, landmarks, score = detect_and_align.align_image(
frame, self.pnet, self.rnet, self.onet)
# if found faces
if len(face_patches) > 0:
self.shared_variables.detection_score[self.index] = score
self.no_face_count = 0
# Save landmark
#self.shared_variables.landmarks[self.index] = self.convert_to_dlib_landmarks(landmarks)
self.shared_variables.set_landmarks(
self.convert_to_dlib_landmarks(landmarks), self.index)
# Convert box from Tensorflow to OpenCV
i = 0
while i < len(padded_bounding_boxes):
face_box.append(
self.convert_tensorflow_box_to_openCV_box(
padded_bounding_boxes[i]))
i += 1
# Save boxes
self.shared_variables.face_box[self.index] = face_box
self.shared_variables.set_detection_box(
face_box, self.index)
#self.shared_variables.detection_box[self.index] = face_box
self.shared_variables.face_found[self.index] = True
# Do flipp test on detection
if self.shared_variables.flipp_test[
self.index] and self.do_flipp_test:
# save flipp as success
degree = self.shared_variables.flipp_test_degree[
self.
index] + self.flipp_test_nr * self.flipp_test_degree
degree = degree - (degree % 360) * 360
self.shared_variables.flipp_test_degree[
self.index] = degree
# log frame change
LOG.info(
"Flipp test successful add degree :" +
str(self.flipp_test_nr * self.flipp_test_degree),
self.shared_variables.name)
# end flipp test
self.do_flipp_test = False
self.flipp_test_nr = 1
# Wake tracking thread
# if not self.shared_variables.tracking_running:
# self.sleep_time = self.SHORT_SLEEP
else:
# No face
self.shared_variables.face_found[self.index] = False
# if max face misses has been done, stop tracking and do less detections
if self.no_face_count >= self.NO_FACE_MAX:
# do flipp test
if self.shared_variables.flipp_test[self.index]:
# doing flipp test
if self.do_flipp_test:
self.flipp_test_nr = self.flipp_test_nr + 1
# flipp test did not find anything
if self.flipp_test_nr * self.flipp_test_degree >= 360:
self.do_flipp_test = False
self.flipp_test_nr = 1
#self.sleep_time = self.LONG_SLEEP
#self.shared_variables.tracking_running = False
#LOG.log("Initiate energy save",self.shared_variables.name)
else:
self.do_flipp_test = True
else:
#self.sleep_time = self.LONG_SLEEP
#self.shared_variables.tracking_running = False
#LOG.log("Initiate energy save",self.shared_variables.name)
pass
else:
self.no_face_count = self.no_face_count + 1
if self.no_face_count >= self.flipp_test_long_intervall and self.shared_variables.flipp_test[
self.index]:
self.no_face_count = 0
self.end_time = datetime.datetime.now()
# Debug detection time
if self.shared_variables.debug:
LOG.debug(
'TF Detection time:' +
str(self.end_time - self.start_time),
self.shared_variables.name)
time.sleep(self.sleep_time) # sleep if wanted
LOG.info('Ending tf detection' + str(self.index),
"SYSTEM-" + self.shared_variables.name)
def run(self):
# Load model
LOG.info("Loading OPENCV model" + str(self.index),
"SYSTEM-" + self.shared_variables.name)
face_cascade = cv2.CascadeClassifier(
'utils/haarcascade_frontalface_default.xml')
facial_features_cascade = cv2.CascadeClassifier(
'utils/haarcascade_facial_features.xml')
LOG.info("Start opencv detections " + str(self.index),
"SYSTEM-" + self.shared_variables.name)
# Start Loop
while self.shared_variables.system_running:
self.start_time = datetime.datetime.now()
frame = self.shared_variables.frame[self.index]
if self.do_flipp_test:
frame = imutils.rotate(
frame, self.flipp_test_degree * self.flipp_test_nr)
# Do detection
if frame is not None:
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
landmarksAndFaces = []
face_patches = face_cascade.detectMultiScale(gray, 1.3, 5)
# if found faces
if len(face_patches) > 0:
landmarksAndFaces.append(face_patches[0].tolist())
for (x, y, w, h) in face_patches:
roi_gray = gray[y:y + h, x:x + w]
# To dont use landmarks, instead use boxes
for (ex, ey, ew,
eh) in facial_features_cascade.detectMultiScale(
roi_gray):
landmarksAndFaces.append([x + ex, y + ey, ew, eh])
self.no_face_count = 0
self.shared_variables.face_found[self.index] = True
# Save boxes
self.shared_variables.face_box[
self.index] = landmarksAndFaces
#self.shared_variables.detection_box[self.index] = face_box
self.shared_variables.set_detection_box(
landmarksAndFaces, self.index)
# Do flipp test on detection
if self.shared_variables.flipp_test[
self.index] and self.do_flipp_test:
# save flipp as success
degree = self.shared_variables.flipp_test_degree[
self.
index] + self.flipp_test_nr * self.flipp_test_degree
degree = degree - (degree % 360) * 360
self.shared_variables.flipp_test_degree[
self.index] = degree
# log frame change
LOG.log(
"Flipp test successful add degree :" +
str(self.flipp_test_nr * self.flipp_test_degree),
self.shared_variables.name)
# end flipp test
self.do_flipp_test = False
self.flipp_test_nr = 1
else:
# No face
self.shared_variables.face_found[self.index] = False
# if max face misses has been done, stop tracking and do less detections
if self.no_face_count >= self.NO_FACE_MAX:
# do flipp test
if self.shared_variables.flipp_test:
# doing flipp test
if self.do_flipp_test:
self.flipp_test_nr = self.flipp_test_nr + 1
# flipp test did not find anything
if self.flipp_test_nr * self.flipp_test_degree >= 360:
self.do_flipp_test = False
self.flipp_test_nr = 1
self.sleep_time = self.LONG_SLEEP
else:
self.do_flipp_test = True
else:
#self.sleep_time = self.LONG_SLEEP
#self.shared_variables.tracking_running = False
#LOG.log("Initiate energy save",self.shared_variables.name)
pass
else:
self.no_face_count = self.no_face_count + 1
if self.no_face_count >= self.flipp_test_long_intervall and self.shared_variables.flipp_test[
self.index]:
self.no_face_count = 0
self.end_time = datetime.datetime.now()
# Debug detection time
if self.shared_variables.debug:
LOG.debug(
'OPENCV Detection time:' +
str(self.end_time - self.start_time),
self.shared_variables.name)
time.sleep(self.sleep_time) # sleep if wanted
LOG.info("Ending OPENCV detection " + str(self.index),
"SYSTEM-" + self.shared_variables.name)
def run(self):
if not self.Loaded_model:
LOG.info("Loading Dlib modell" + str(self.index),
"SYSTEM-" + self.shared_variables.name)
# Load model
self.face_detector = dlib.get_frontal_face_detector()
self.landmarks_predictor = dlib.shape_predictor(
self.landmarks_model_path)
#face_cascade = cv2.CascadeClassifier(face_cascade_path)
self.Loaded_model = True
LOG.info("Start dlib detections" + str(self.index),
"SYSTEM-" + self.shared_variables.name)
#wait for first cam frame
while self.shared_variables.frame[self.index] is None:
pass
# Start Loop
while self.shared_variables.system_running:
self.start_time = datetime.datetime.now()
frame = self.shared_variables.frame[self.index]
if self.do_flipp_test:
frame = imutils.rotate(
frame, self.flipp_test_degree * self.flipp_test_nr)
# Do detection
success, face_box, landmarks, score = self.object_detection(frame)
# if found faces
if success:
self.shared_variables.detection_score[self.index] = score
self.no_face_count = 0
# Save landmark
#self.shared_variables.landmarks[self.index] = landmarks
self.shared_variables.set_landmarks(landmarks, self.index)
# Save boxes
self.shared_variables.face_box[self.index] = face_box
#self.shared_variables.detection_box[self.index] = face_box
self.shared_variables.set_detection_box(face_box, self.index)
self.shared_variables.face_found[self.index] = True
# Do flipp test on detection
if self.shared_variables.flipp_test[
self.index] and self.do_flipp_test:
# save flipp as success
degree = self.shared_variables.flipp_test_degree[
self.
index] + self.flipp_test_nr * self.flipp_test_degree
degree = degree - (degree % 360) * 360
self.shared_variables.flipp_test_degree[
self.index] = degree
# log frame change
LOG.info(
"Flipp test successful add degree :" +
str(self.flipp_test_nr * self.flipp_test_degree),
self.shared_variables.name)
# end flipp test
self.do_flipp_test = False
self.flipp_test_nr = 1
# Wake tracking thread
#if not self.shared_variables.tracking_running[self.index]:
# self.sleep_time = self.SHORT_SLEEP
else:
# No face
self.shared_variables.face_found[self.index] = False
# if max face misses has been done, do less detections
if self.no_face_count >= self.NO_FACE_MAX:
# do flipp test
if self.shared_variables.flipp_test[self.index]:
# doing flipp test
if self.do_flipp_test:
self.flipp_test_nr = self.flipp_test_nr + 1
# flipp test did not find anything
if self.flipp_test_nr * self.flipp_test_degree >= 360:
self.do_flipp_test = False
self.flipp_test_nr = 1
if self.sleep_time == self.SHORT_SLEEP:
#LOG.log("Initiate energy save",self.shared_variables.name)
#self.sleep_time = self.LONG_SLEEP
pass
else:
self.do_flipp_test = True
else:
#self.sleep_time = self.LONG_SLEEP
#self.shared_variables.tracking_running[self.index] = False
#LOG.log("Initiate energy save",self.shared_variables.name)
pass
else:
self.no_face_count = self.no_face_count + 1
if self.no_face_count >= self.flipp_test_long_intervall and self.shared_variables.flipp_test[
self.index]:
self.no_face_count = 0
self.end_time = datetime.datetime.now()
# Debug detection time
if self.shared_variables.debug:
LOG.debug(
'Dlib Detection time:' +
str(self.end_time - self.start_time),
self.shared_variables.name)
time.sleep(self.sleep_time) # sleep if wanted
LOG.info("Ending dlib detection " + str(self.index),
"SYSTEM-" + self.shared_variables.name)
def run(self):
LOG.info("Start age and gender estimation "+ str(self.index), "SYSTEM-"+self.shared_variables.name)
# load model and weights
img_size = 64
model = WideResNet(img_size, depth=16, k=8)()
model.load_weights(self.pretrained_model_path)
#wait for detection
while self.shared_variables.frame[self.index] is None:
pass
img_size = 64
while self.shared_variables.system_running:
input_img = cv2.cvtColor(self.shared_variables.frame[self.index], cv2.COLOR_BGR2RGB)
img_h, img_w, _ = np.shape(input_img)
faces = np.empty((len([self.shared_variables.face_box[self.index]]), img_size, img_size, 3))
w = self.shared_variables.face_box[self.index][0][0]
h = self.shared_variables.face_box[self.index][0][1]
x1 = self.shared_variables.face_box[self.index][0][2]
y1 = self.shared_variables.face_box[self.index][0][3]
x2 = w + x1
y2 = h + y1
xw1 = max(int(x1 - 0.4 * w), 0)
yw1 = max(int(y1 - 0.4 * h), 0)
xw2 = min(int(x2 + 0.4 * w), img_w - 1)
yw2 = min(int(y2 + 0.4 * h), img_h - 1)
faces[0, :, :, :] = cv2.resize(self.shared_variables.frame[self.index][yw1:yw2 + 1, xw1:xw2 + 1, :], (img_size, img_size))
# predict ages and genders of the detected faces
results = model.predict(faces)
predicted_genders = results[0]
ages = np.arange(0, 101).reshape(101, 1)
predicted_ages = results[1].dot(ages).flatten()
# Show
if(self.shared_variables.debug):
LOG.debug("Predicted age: " + str(predicted_ages[0]) + " from " + str(self.index), "SYSTEM-"+self.shared_variables.name)
gender = ""
if predicted_genders[0][0] > 0.5:
gender = "Female"
else:
gender = "Male"
if(self.shared_variables.debug):
print("Predicted gender: " + gender)
self.shared_variables.gender[self.index] = gender
self.shared_variables.age[self.index] = predicted_ages[0]
#Short version
if(self.shared_variables.debug):
label = "{}, {}".format(int(predicted_ages[0]),
"F" if predicted_genders[0][0] > 0.5 else "M")
LOG.debug(label +" from " + str(self.index), "SYSTEM-"+self.shared_variables.name)
LOG.info("Stopped age gender estimation" + str(self.index), "SYSTEM-"+self.shared_variables.name)