Ask Your Question

Python. Simple image Face Detection using OpenCV

asked 2013-07-07 03:46:26 -0600

CharlyChi gravatar image

I'm noobe in OpenCV and looking for simple python face detecting script. I've find one here

But there is no rect. on face when script stops.


import cv2
img = cv2.imread("Lenna.png")
hc = cv2.CascadeClassifier("haarcascade_frontalface_alt2.xml")
faces = hc.detectMultiScale(img)

for face in faces:
  cv2.rectangle(img, (face[0], face[1]), (face[0] + face[2], face[0] + face[3]), (255, 0, 0), 3)

cv2.imshow("Lenna's face", img)
if cv2.waitKey(5000) == 27:
  cv2.destroyWindow("Lenna's face")
cv2.imwrite("LennaFace.png", img)
edit retag flag offensive close merge delete


That should be cv2.rectangle(img, (face[0], face[1]), (face[0] + face[2], face[1] + face[3]), (255, 0, 0), 3) -- not face[0] + face[3]

Wilfred Hughes gravatar imageWilfred Hughes ( 2013-09-20 10:59:23 -0600 )edit

face[0] + face[3] at line 8 should be face[1] + face[3] . It also makes sense to iterate over tuples of (x, y, w, h) and set points to (x, y) and (x + w, y + h).

ffriend gravatar imageffriend ( 2013-10-19 15:35:05 -0600 )edit

2 answers

Sort by ยป oldest newest most voted

answered 2013-11-02 03:27:48 -0600

There's no needed for grayscale images. Only thing you needed to do is, be sure "haarcascade_frontalface_alt2.xml" file nearby your source code directory. If not write the whole path to it.

Ps.: you may cheat from .../opencv-(version)/data/haarcascades/haarcascade_frontalface_alt2.xml

edit flag offensive delete link more

answered 2013-07-07 04:49:43 -0600

berak gravatar image

you need a grayscale image for this.

could convert after loading, but telling imread to do it, is easier:

img = cv2.imread("Lenna.png",0); // load 8bit gray
edit flag offensive delete link more


Works fine in my testing without converting to grayscale.

Wilfred Hughes gravatar imageWilfred Hughes ( 2013-09-20 10:58:38 -0600 )edit
Login/Signup to Answer

Question Tools


Asked: 2013-07-07 03:46:26 -0600

Seen: 2,206 times

Last updated: Nov 02 '13