def webcam():
if request.method == 'GET':
return render_template('upload.html')
elif request.method == 'POST':
image_b64 = request.form['img']
print(type(image_b64))
robo_talker.talker(image_b64)
# 等待返回结果
time.sleep(1)
result = ''
is_fatigue = ''
mar = ''
ear = ''
if robo_talker.f_img != None:
fat_result_path = robo_talker.f_img
is_fatigue, mar, ear = fat_result_path.split(',')
robo_talker.f_img = None
if robo_talker.drive_state != None:
result = robo_talker.drive_state
robo_talker.drive_state = None
return 'state: {0}, is_fatigue: {1}, mar: {2}, ear: {3}'.format(
result, is_fatigue, mar, ear)
def upload():
print('getting data from web.')
if request.method == 'POST':
image_b64 = request.form['img']
robo_talker.talker(image_b64)
imgdata = base64.b64decode(image_b64)
with open('org_img.jpg', 'wb') as f:
f.write(imgdata)
imgdata = cv2.imread('org_img.jpg')
imgdata = cv2.cvtColor(imgdata, cv2.COLOR_BGR2GRAY)
face_cascade = cv2.CascadeClassifier(
r'./haarcascade_frontalface_default.xml')
faces = face_cascade.detectMultiScale(
imgdata,
scaleFactor=1.15,
minNeighbors=5,
minSize=(5, 5) # ,
# flags=cv2.CV_HAAR_SCALE_IMAGE
)
for i, (x, y, w, h) in enumerate(faces):
cv2.rectangle(imgdata, (x, y), (x + w, y + w), (255, 0, 0), 2)
gray_face = imgdata[y:y + h, x:x + w]
gray_face_vector = extract_face_features(gray_face)
gray_face = gray_face_vector.reshape((96, 96))
cv2.imwrite('./gray_face.jpg', gray_face)
time.sleep(3)
if robo_talker.MSG != None:
result = robo_talker.MSG
robo_talker.MSG = None
return "result:" + result
else:
return "无法识别"
def upload():
print('getting data from web.')
if request.method == 'POST':
image_b64 = request.form['img']
robo_talker.talker(image_b64)
imgdata = base64.b64decode(image_b64)
with open('org_img.jpg', 'wb') as f:
f.write(imgdata)
time.sleep(3)
if robo_talker.MSG != None:
result = robo_talker.MSG
robo_talker.MSG = None
return "result:" + result
else:
return "无法识别"
cv2.imshow('frame',frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
else:
break
# Release everything if job is finished
cap.release()
out.release()
cv2.destroyAllWindows()
# start waiting result
print("waiting...")
message = String()
message.data = "video detection start"
robo_talker.talker(message)
# then check the result
if robo_talker.MSG != None:
result = robo_talker.MSG
robo_talker.MSG = None
print("result:" + result)
else:
print("无法识别")
'''
videoCapture = cv2.VideoCapture()
videoCapture.open('output.avi')
fps = videoCapture.get(cv2.CAP_PROP_FPS)
def upload():
print('getting data from web.')
if request.method == 'POST':
image_b64 = request.form['img']
robo_talker.talker(image_b64)
imgdata = base64.b64decode(image_b64)
with open('img/org_img.jpg', 'wb') as f:
f.write(imgdata)
# imgdata = cv2.imread('img/org_img.jpg')
# imgdata = cv2.cvtColor(imgdata, cv2.COLOR_BGR2GRAY)
# face_cascade = cv2.CascadeClassifier(r'./haarcascade_frontalface_default.xml')
# faces = face_cascade.detectMultiScale(
# imgdata,
# scaleFactor=1.15,
# minNeighbors=5,
# minSize=(5, 5) # ,
# # flags=cv2.CV_HAAR_SCALE_IMAGE
# )
# for i, (x, y, w, h) in enumerate(faces):
# cv2.rectangle(imgdata, (x, y), (x + w, y + w), (255, 0, 0), 2)
# gray_face = imgdata[y:y+h, x:x+w]
# gray_face_vector = extract_face_features(gray_face)
# gray_face = gray_face_vector.reshape((96, 96))
# cv2.imwrite('./gray_face.jpg', gray_face)
global db_conn
time.sleep(5)
if robo_talker.MSG != None:
result = robo_talker.MSG
robo_talker.MSG = None
#DB_Handler
sql = "select * from Cars where licence = %s"
car = None
with db_conn.cursor() as cursor:
cursor.execute(sql, (result))
car = cursor.fetchone()
db_conn.commit()
print(car)
if (car == None):
cur_time = datetime.datetime.now().strftime(
'%Y-%m-%d %H:%M:%S')
print(cur_time)
sql = "insert into Cars (licence, time) values(%s, %s);"
with db_conn.cursor() as cursor:
cursor.execute(sql, (result, cur_time))
db_conn.commit()
return "Welcome~ " + result + "\nEnter at " + cur_time
else:
park_time = datetime.datetime.strptime(car['time'],
'%Y-%m-%d %H:%M:%S')
cur_time = datetime.datetime.now()
sql = "delete from Cars where licence = %s"
with db_conn.cursor() as cursor:
cursor.execute(sql, (result))
db_conn.commit()
return result + " has parked for " + str(
round((cur_time - park_time).seconds / 60, 2)) + " minutes"
else:
return "Can't recognize"