fonfonx fonfonx - 2 years ago 1192
Python Question

OpenCV detect face landmarks (ear-chin-ear line)

I am looking to an opencv function (in python) detecting the line left ear - chin - right ear (that looks like a parabol) on human faces. Is there any kind of haarcascade doing this job? I already know the frontal face or the eyes haarcascades but I am looking for something more precise.

Answer Source

what you are looking for is called face landmark detection . You can try DLIB . DLIB is written in C++ but it also have a python wrapper .Install Instructions

Now using DLib you can achieve this


import cv2
import dlib
import numpy

PREDICTOR_PATH = "/home/zed/dlib/files/shape_predictor_68_face_landmarks.dat"
predictor = dlib.shape_predictor(PREDICTOR_PATH)
cascade = cv2.CascadeClassifier(cascade_path)

# #This is using the Dlib Face Detector . Better result more time taking
# def get_landmarks(im):
#     rects = detector(im, 1)
#     rect=rects[0]
#     print type(rect.width())
#     fwd=int(rect.width())
#     if len(rects) == 0:
#         return None,None

#     return np.matrix([[p.x, p.y] for p in predictor(im, rects[0]).parts()]),fwd

def get_landmarks(im):
    rects = cascade.detectMultiScale(im, 1.3,5)
    x,y,w,h =rects[0]
    return numpy.matrix([[p.x, p.y] for p in predictor(im, rect).parts()])

def annotate_landmarks(im, landmarks):
    im = im.copy()
    for idx, point in enumerate(landmarks):
        pos = (point[0, 0], point[0, 1])
        cv2.putText(im, str(idx), pos,
                    color=(0, 0, 255)), pos, 3, color=(0, 255, 255))
    return im



enter image description here

Points related to landmark :

FACE_POINTS = list(range(17, 68))
MOUTH_POINTS = list(range(48, 61))
RIGHT_BROW_POINTS = list(range(17, 22))
LEFT_BROW_POINTS = list(range(22, 27))
RIGHT_EYE_POINTS = list(range(36, 42))
LEFT_EYE_POINTS = list(range(42, 48))
NOSE_POINTS = list(range(27, 35))
JAW_POINTS = list(range(0, 17))
Recommended from our users: Dynamic Network Monitoring from WhatsUp Gold from IPSwitch. Free Download