def shutdown(self, signum):
MMConfig.toNode("payload", {
"type": "status",
"status": "Shutdown: Cleaning up camera..."
})
camera.stop()
quit()
def detectMotion(image):
global t_minus, t, t_plus
global last_motion
t_minus = t
t = t_plus
t_plus = cv2.cvtColor(camera.read(), cv2.COLOR_RGB2GRAY)
diff = diffImg(t_minus, t, t_plus)
thresh = cv2.threshold(diff, 25, 255, cv2.THRESH_BINARY)[1]
# dilate the thresholded image to fill in holes, then find contours
# on thresholded image
thresh = cv2.dilate(thresh, None, iterations=2)
(_, cnts, _) = cv2.findContours(thresh.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
max = 0
# loop over the contours
for c in cnts:
area = cv2.contourArea(c)
if area > max:
max = area
if max > MMConfig.getMotionDetectionThreshold():
if last_motion is None:
MMConfig.toNode("motion-detected", {})
last_motion = time.time()
elif last_motion != None and time.time() - last_motion > MMConfig.getMotionStopDelay():
last_motion = None
MMConfig.toNode("motion-stopped", {})
def detectFace(image):
global current_user, login_timestamp, last_match, same_user_detected_in_row
# Convert image to grayscale.
image = cv2.cvtColor(image, cv2.COLOR_RGB2GRAY)
# Get coordinates of single face in captured image.
result = face.detect_single(image)
# No face found, logout user?
if result is None:
# if last detection exceeds timeout and there is someone logged in -> logout!
if (current_user is not None and time.time() - login_timestamp > MMConfig.getLogoutDelay()):
# callback logout to node helper
MMConfig.toNode("logout", {"user": current_user})
same_user_detected_in_row = 0
current_user = None
return
# Set x,y coordinates, height and width from face detection result
x, y, w, h = result
# Crop image on face. If algorithm is not LBPH also resize because in all other algorithms image resolution has to be the same as training image resolution.
crop = face.crop(image, x, y, w, h, int(MMConfig.getFaceFactor() * w))
# Test face against model.
label, confidence = model.predict(crop)
# We have a match if the label is not "-1" which equals unknown because of exceeded threshold and is not "0" which are negtive training images (see training folder).
if (label != -1 and label != 0):
# Set login time
login_timestamp = time.time()
# Routine to count how many times the same user is detected
if (label == last_match and same_user_detected_in_row < 2):
# if same user as last time increment same_user_detected_in_row +1
same_user_detected_in_row += 1
if label != last_match:
# if the user is diffrent reset same_user_detected_in_row back to 0
same_user_detected_in_row = 0
# A user only gets logged in if he is predicted twice in a row minimizing prediction errors.
if (label != current_user and same_user_detected_in_row > 1):
current_user = label
# Callback current user to node helper
MMConfig.toNode("login", {"user": label, "confidence": str(confidence)})
# set last_match to current prediction
last_match = label
# if label is -1 or 0, current_user is not already set to unknown and last prediction match was at least 5 seconds ago
# (to prevent unknown detection of a known user if he moves for example and can't be detected correctly)
elif (current_user != 0 and time.time() - login_timestamp > 5):
# Set login time
login_timestamp = time.time()
# set current_user to unknown
current_user = 0
# callback to node helper
MMConfig.toNode("login", {"user": current_user, "confidence": None})
def shutdown(self, signum):
MMConfig.toNode("status", 'Shutdown: Cleaning up camera...')
camera.stop()
quit()
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
"""
import sys
import os
sys.path.append((os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))+ '/common/'))
import time
from face import FaceDetection
import cv2
from config import MMConfig
import signal
MMConfig.toNode("status", "Facerecognition started...")
# Setup variables
current_user = None
last_match = None
detection_active = True
login_timestamp = time.time()
same_user_detected_in_row = 0
try:
# Load training data into model
MMConfig.toNode("status", 'Loading training data...')
# load the model
model = cv2.face.LBPHFaceRecognizer_create(threshold=MMConfig.getThreshold())
Reference link - https://github.com/normyx/MMM-Facial-Recognition-OCV3
"""
import sys
import os
sys.path.append(
(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')) +
'/common/'))
import time
from face import FaceDetection
import cv2
from config import MMConfig
import signal
MMConfig.toNode("payload", {
"type": "status",
"status": "Face Detection started..."
})
# Setup variables
isDetected = False
detection_active = True
detected_timestamp = time.time()
# get camera
camera = MMConfig.getCamera()
def shutdown(self, signum):
MMConfig.toNode("payload", {
"type": "status",
"status": "Shutdown: Cleaning up camera..."
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
"""
import sys
import os
sys.path.append((os.path.abspath(os.path.join(os.path.dirname(__file__), '..')) + '/common/'))
import time
from face import FaceDetection
import cv2
from config import MMConfig
import signal
MMConfig.toNode("status", "Facerecognition started...")
# Setup variables
current_user = None
last_match = None
detection_active = True
motion_detection_active = True
login_timestamp = time.time()
same_user_detected_in_row = 0
# Load training data into model
MMConfig.toNode("status", 'Loading training data...')
# load the model
model = cv2.face.LBPHFaceRecognizer_create(threshold=MMConfig.getThreshold())