def get_gender(image: numpy.ndarray):
padding = 20
try:
face_block = FBHelper.get_face_block(image)
if face_block:
startX, startY = max(0,
face_block.left() - padding), max(
0,
face_block.top() - padding)
endX, endY = min(image.shape[1] - 1,
face_block.right() + padding), min(
image.shape[0] - 1,
face_block.bottom() + padding)
face_grid = numpy.copy(image[startY:endY, startX:endX])
label, confidence = cvlib.detect_gender(face_grid)
idx = numpy.argmax(confidence)
if label[idx] == 'male':
return 'Male'
else:
return 'Female'
except Exception as e:
print('Failed to get gender from CVLib. ERROR: {}'.format(e))
def detect_gender(self, image, face_coords):
x0, x1 = face_coords[0:3:2]
y0, y1 = face_coords[1:4:2]
face_crop = np.copy(image[y0:y1, x0:x1])
label, confidence = cv.detect_gender(face_crop)
index = np.argmax(confidence)
gender = label[index]
return gender
def checker(link):
result = 0
img = io.imread(link, plugin='matplotlib')
img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
font = cv2.FONT_HERSHEY_PLAIN
faces, conf = cvlib.detect_face(img)
if len(faces) > 0:
for x, y, w, h in faces:
roi = img[y:y + h, x:x + w]
label, confidence = cvlib.detect_gender(roi)
if confidence[0] > confidence[1]:
gender = label[0]
else:
gender = label[1]
cv2.rectangle(img, (x, y), (w, h), (0, 255, 0), 2)
cv2.rectangle(img, (50, 50), (220, 80), (0, 255, 0), -1)
cv2.putText(img, "Face Detected [+]", (55, 70), font, 1, (0, 0, 0),
1, cv2.LINE_AA)
if gender == "female":
cv2.rectangle(img, (50, 90), (220, 120), (0, 255, 0), -1)
cv2.putText(img, gender + " [+]", (55, 110), font, 1,
(0, 0, 0), 1, cv2.LINE_AA)
result = 1
cv2.rectangle(img, (50, 130), (220, 160), (0, 255, 0), -1)
cv2.putText(img, " Like [+] ", (55, 150), font, 1, (0, 0, 0),
1, cv2.LINE_AA)
else:
cv2.rectangle(img, (50, 90), (220, 120), (0, 0, 255), -1)
cv2.putText(img, gender + " [+]", (55, 110), font, 1,
(0, 0, 0), 1, cv2.LINE_AA)
elif len(faces) == 0:
cv2.rectangle(img, (50, 50), (220, 80), (0, 0, 255), -1)
cv2.putText(img, "Face Not Detected", (55, 70), font, 1,
(255, 255, 255), 1, cv2.LINE_AA)
cv2.rectangle(img, (50, 90), (220, 120), (0, 0, 255), -1)
cv2.putText(img, "Dislike [-]", (55, 110), font, 1, (255, 255, 255), 1,
cv2.LINE_AA)
cv2.imshow('Tinder Detector', img)
cv2.waitKey(2000)
cv2.destroyAllWindows()
return result
def genderDectection(img_path, output_folder):
#Creating output folder if not present
if output_folder and not os.path.isdir(output_folder):
os.mkdir(output_folder)
# read input image
img = cv2.imread(img_path)
# apply face detection
face, conf = cv.detect_face(img)
padding = 5
# loop through detected faces
for f in face:
(startX, startY) = max(0, f[0] - padding), max(0, f[1] - padding)
(endX, endY) = min(img.shape[1] - 1,
f[2] + padding), min(img.shape[0] - 1,
f[3] + padding)
# draw rectangle over face
cv2.rectangle(img, (startX, startY), (endX, endY), (0, 255, 0), 2)
face_crop = np.copy(img[startY:endY, startX:endX])
# apply gender detection
(label, confidence) = cv.detect_gender(face_crop)
idx = np.argmax(confidence)
label = label[idx]
label = "{}: {:.2f}%".format(label, confidence[idx] * 100)
Y = startY - 5 if startY - 5 > 5 else startY + 5
cv2.putText(img, label, (startX, Y), cv2.FONT_HERSHEY_SIMPLEX, 0.6,
(0, 255, 255), 1)
# converting back to RBG
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
#Save the image to output folder
if output_folder:
input_filename, ext = os.path.splitext(os.path.basename(img_path))
filename = os.path.join(output_folder, input_filename.format(ext))
fig = plt.figure()
plt.imshow(img)
plt.axis("off")
plt.savefig(filename, bbox_inches='tight', pad_inches=0)
plt.close(fig)
def gender_detect(img):
image = cv2.imread(img)
face, conf = cv.detect_face(image)
padding = 20
for f in face:
(startX, startY) = max(0, f[0] - padding), max(0, f[1] - padding)
(endX, endY) = min(image.shape[1] - 1,
f[2] + padding), min(image.shape[0] - 1,
f[3] + padding)
cv2.rectangle(image, (startX, startY), (endX, endY), (0, 255, 0), 2)
face_crop = np.copy(image[startY:endY, startX:endX])
(label, confidence) = cv.detect_gender(face_crop)
idx = np.argmax(confidence)
conf = np.max(confidence).item()
label = label[idx]
return {"conf": conf, "label": label}
def detect(self, f,ext, args):
fi = f+ext
fo = f+'-face'+ext
g.log.debug("Reading {}".format(fi))
image = cv2.imread(fi)
faces, conf = cv.detect_face(image)
detections = []
for f, c in zip(faces, conf):
c = "{:.2f}%".format(c * 100)
(startX, startY) = f[0], f[1]
(endX, endY) = f[2], f[3]
cv2.rectangle(image, (startX, startY),
(endX, endY), (0, 255, 0), 2)
rect = [int(startX), int(startY), int(endX), int(endY)]
obj = {
'type': 'face',
'confidence': c,
'box': rect
}
if args['gender']:
face_crop = np.copy(image[startY:endY, startX:endX])
(gender_label_arr, gender_confidence_arr) = cv.detect_gender(face_crop)
idx = np.argmax(gender_confidence_arr)
gender_label = gender_label_arr[idx]
gender_confidence = "{:.2f}%".format(
gender_confidence_arr[idx] * 100)
obj['gender'] = gender_label
obj['gender_confidence'] = gender_confidence
Y = startY - 10 if startY - 10 > 10 else startY + 10
cv2.putText(image, gender_label, (startX, Y),
cv2.FONT_HERSHEY_SIMPLEX, 0.7, (0, 255, 0), 2)
detections.append(obj)
if not args['delete']:
g.log.debug("Writing {}".format(fo))
cv2.imwrite(fo, image)
if args['delete']:
os.remove(fi)
return detections
def imageRecog(url):
if url.startswith('https'):
image = url_to_image(url)
else:
image = cv2.imread(url)
image = cv2.cvtColor(image, cv2.IMREAD_COLOR)
bbox, label, conf = cv.detect_common_objects(image)
list_of_things = []
list_of_genders = []
list_of_gender_conf = []
personInfo = []
for i in range(len(label)):
# object = object.lstrip("['")
# object = object.rstrip("']")
# confidence = confidence.lstrip("['")
# confidence = confidence.rstrip("']")
if label[i] not in list_of_things:
list_of_things.append(label[i])
if label[i] == 'person':
label2, confidence = cv.detect_gender(image, False)
if confidence[0] > confidence[1]: #Male
confidence[0] = int(confidence[0] * 100)
list_of_genders.append(confidence[0])
list_of_gender_conf.append(label2[0])
personInfo = [list_of_genders, list_of_gender_conf]
else:
confidence[1] = int(confidence[1] * 100)
list_of_genders.append(confidence[1])
list_of_gender_conf.append(label2[1])
personInfo = [list_of_genders, list_of_gender_conf]
faces, confidences = cv.detect_face(image)
confidence = 0
for i in range(len(conf)):
confidence = confidence + conf[i]
avgConf = confidence / len(conf)
return list_of_things, avgConf, personInfo
def pose_estimation(IMAGE, flag):
thr = 0.1
inWidth = 368
inHeight = 368
net = cv.dnn.readNetFromTensorflow("graph_opt.pb")
cap = cv.VideoCapture(IMAGE)
cor = color()
font = cv.FONT_HERSHEY_SCRIPT_SIMPLEX
frame = None
status = {'pose': 0}
while cv.waitKey(1) < 0:
hasFrame, frame = cap.read()
try:
face, conf = cvlib.detect_face(frame)
padding = 20
for f in face:
(startX, startY) = max(0,
f[0] - padding), max(0, f[1] - padding)
(endX,
endY) = min(frame.shape[1] - 1,
f[2] + padding), min(frame.shape[0] - 1,
f[3] + padding)
cv.rectangle(frame, (startX, startY), (endX, endY),
(0, 255, 0), 2)
(label, confidence) = cvlib.detect_gender(
np.copy(frame[startY:endY, startX:endX]))
idx = np.argmax(confidence)
label = label[idx]
conf = confidence[idx] * 100
if conf > 90:
cv.putText(frame, "{}: {:.2f}%".format(label, conf),
(startX, startY -
10 if startY - 10 > 10 else startY + 10),
cv.FONT_HERSHEY_SIMPLEX, 0.7, (0, 255, 0), 2)
if label == 'male' and flag == True:
status['pose'] = 'aprovado'
else:
status['pose'] = 'reprovado'
else:
print("...")
except TypeError:
pass
if not hasFrame:
cv.waitKey()
break
frameWidth = frame.shape[1]
frameHeight = frame.shape[0]
net.setInput(
cv.dnn.blobFromImage(frame,
1.0, (inWidth, inHeight),
(127.5, 127.5, 127.5),
swapRB=True,
crop=False))
out = net.forward()
out = out[:, :19, :, :]
assert (len(BODY_PARTS) == out.shape[1])
points = []
for i in range(len(BODY_PARTS)):
heatMap = out[0, i, :, :]
_, conf, _, point = cv.minMaxLoc(heatMap)
points.append((int((frameWidth * point[0]) / out.shape[3]),
int((frameHeight * point[1]) /
out.shape[2])) if conf > thr else None)
for pair in POSE_PAIRS:
partFrom = pair[0]
partTo = pair[1]
assert (partFrom in BODY_PARTS)
assert (partTo in BODY_PARTS)
idFrom = BODY_PARTS[partFrom]
idTo = BODY_PARTS[partTo]
if points[idFrom] and points[idTo]:
cv.line(frame, points[idFrom], points[idTo], (cor, cor, cor),
3)
print(idFrom, idTo, points[idFrom], points[idTo])
cv.ellipse(frame, points[idFrom], (4, 4), 0, 0, 360,
(cor, cor, cor), cv.FILLED)
cv.ellipse(frame, points[idTo], (4, 4), 0, 0, 360,
(cor, cor, cor), cv.FILLED)
t, _ = net.getPerfProfile()
freq = cv.getTickFrequency() / 1000
cv.putText(frame, '%.2fms' % (t / freq), (10, 20),
cv.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 0))
#cv.imshow('OpenPose using OpenCV', frame)
name = "output_" + str(uuid4()) + IMAGE
cv.imwrite(name, frame)
return status
import numpy as np
webcam = cv2.VideoCapture(0)
padding = 20
while webcam.isOpened():
status, frame = webcam.read()
face, confidence = cv.detect_face(frame)
for idx, f in enumerate(face):
(startX, startY) = max(0, f[0] - padding), max(0, f[1] - padding)
(endX, endY) = min(frame.shape[1] - 1,
f[2] + padding), min(frame.shape[0] - 1,
f[3] + padding)
cv2.rectangle(frame, (startX, startY), (endX, endY), (0, 255, 0), 2)
face_crop = np.copy(frame[startY:endY, startX:endX])
(label, confidence) = cv.detect_gender(face_crop)
idx = np.argmax(confidence)
label = label[idx]
label = "{}: {:.2f}%".format(label, confidence[idx] * 100)
Y = startY - 10 if startY - 10 > 10 else startY + 10
cv2.putText(frame, label, (startX, Y), cv2.FONT_HERSHEY_SIMPLEX, 0.7,
(0, 255, 0), 2)
cv2.imshow("Real-time gender detection", frame)
# press "s" to stop
if cv2.waitKey(1) & 0xFF == ord('s'):
break
webcam.release()
cv2.destroyAllWindows()
# In[ ]:
import glob
students = glob.glob("./Images/*.jpg")
# print(students)
# exit()
n = 0
for pics in students:
try:
image = cv2.imread(pics)
label, confidence = cv.detect_gender(image, enable_gpu=True)
# print(confidence)
idx = np.argmax(confidence)
label = label[idx]
gender_acc = "{}: {:.2f}%".format(label, confidence[idx] * 100)
# cordinates for cv2.imwrite
padding = 10
startX = max(0, (image[0] - padding).all())
startY = max(0, (image[1] - padding).all())
if startY - 10 > 10:
Y = startY - 20
else:
Y = startY + 20
import cvlib as cv
import numpy as np
img = cv2.imread('sample3.png')
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
# 얼굴 찾기
faces, confidences = cv.detect_face(img)
for (x, y, x2, y2)in faces:
# 얼굴 roi 지정
face_img = img[y:y2, x:x2]
# 성별 예측하기
label, confidence = cv.detect_gender(face_img)
cv2.rectangle(img, (x, y), (x2, y2), (0, 255, 0), 2)
gender = np.argmax(confidence)
text = f'{label[gender]}:{confidence[gender]:.1%}'
cv2.putText(img, text, (x,y-10), cv2.FONT_HERSHEY_PLAIN, 1, (0, 255, 0), 1)
# 영상 출력
cv2.imshow('image', img)
key = cv2.waitKey(0)
cv2.destroyAllWindows()
https://github.com/arunponnusamy/cvlib
for idx, f in enumerate(detectedFaces):
# Get rectangle coordinates init XY coordinates and last XY coordinates
(initX,initY) = max(0, f[0] - padding), max(0, f[1] - padding)
(lastX,lastY) = min(rgbFrame.shape[1] - 1, f[2] + padding), min(rgbFrame.shape[0]-1, f[3] + padding)
# Print red rectangle in detected faces
cv2.rectangle(in_buf_array, (initX,initY), (lastX,lastY), (255,0,0), 2)
# Gender analysis
# Extract detected faces
extractedGenderFace = np.copy(rgbFrame[initY:lastY, initX:lastX])
# Detect gender
(genderDetectionResult, genderDetectionConfidence) = cv.detect_gender(extractedGenderFace)
# Prepare label and confidence value
idx = np.argmax(genderDetectionConfidence)
genderDetectionLabel = genderDetectionResult[idx]
genderDetectionLabel = str(genderDetectionLabel) + " " + str(int(genderDetectionConfidence[idx] * 100)) + "%"
# Get detected gender value to send
genderDetection = genderDetectionLabel
# Update full detection label
detectionFrameLabel = "G: " + str(genderDetection)
# Age detection
# Extract detected faces
extractedAgeFace = rgbFrame[initY:lastY, initX:lastX].copy()
