def videoLoop(self):
try:
# keep looping over frames until we are instructed to stop
i = 0
while not self.stopEvent.is_set():
# grab the frame from the video stream and resize it to
# have a maximum width of 300 pixels
flag, self.frame = vs.read()
if i == 0:
i += 2
faceCoordinates = None
else:
try:
faceCoordinates = fd.face_detect(self.frame)
startX = faceCoordinates[0]
startY = faceCoordinates[1]
endX = faceCoordinates[2]
endY = faceCoordinates[3]
if startX is not None:
image = fd.draw_rect(self.frame, startX, startY,
endX, endY)
face_img = fd.face_crop(self.frame,
faceCoordinates,
face_shape=(128, 128))
im = img_to_array(face_img)
im = np.expand_dims(im, axis=0)
result = self.model.predict(im)
for i in range(0, 7):
self.resultLabel[i].config(
text=str(round(result[0][i], 4)))
except:
pass
self.frame = imutils.resize(self.frame, width=300)
self.frame = cv2.flip(self.frame, 1)
# OpenCV represents images in BGR order; however PIL
# represents images in RGB order, so we need to swap
# the channels, then convert to PIL and ImageTk format
image = cv2.cvtColor(self.frame, cv2.COLOR_BGR2RGB)
image = Image.fromarray(image)
image = ImageTk.PhotoImage(image)
# if the panel is not None, we need to initialize it
if self.panelA is None:
self.panelA = tki.Label(image=image)
self.panelA.image = image
self.panelA.pack(side="left", padx=10, pady=10)
self.panelA.place(x=20, y=20)
# otherwise, simply update the panel
else:
self.panelA.configure(image=image)
self.panelA.image = image
except RuntimeError:
print("[INFO] caught a RuntimeError")
def refreshFrame(frame, startX, startY, endX, endY, index):
if startX is not None:
fd.draw_rect(frame, startX, startY, endX, endY, index)
cv2.imshow(windowName, frame)
faces, faceID = fr.labels_for_training_data(
'E:/python_programs/face detection/training')
face_recognizer = fr.train_classifier(faces, faceID)
face_recognizer.save('trainingData.yml')
name = {0: "person1", 1: "person2", 2: "person3", 3: "person4"}
# face_recognizer=cv2.face.LBPHFaceRecognizer_create()
# face_recognizer.read('E:/python_programs/face detection/trainingData.yml')
for face in faces_detected:
(x, y, w, h) = face
roi_gray = gray_img[y:y + w, x:x + h]
label, confidence = face_recognizer.predict(roi_gray)
print("confidence:", confidence)
print("label:", label)
fr.draw_rect(test_img, face)
if (confidence > 100):
continue
predicted_name = name[label]
fr.put_text(test_img, predicted_name, x, y)
# #resize the image inorder to fit the rectangle
resized_img = cv2.resize(test_img,
(1000, 700)) # resizes te image to 1000X700
cv2.imshow("face_detected:", resized_img)
time.sleep(1)
fr.to_audio(name, label)
f = open('name.txt', 'w')
f.write("the person/s is {}".format(name[label]))
f.close()