putting the detected face into an image and display in another window in opencv using python

Question

i'm new to image processing and opencv, but so far the easy to understand functions and good documentation have enabled me to try out and understand upto some level code like facedetection etc.

Now when i detect the faces in the webcam video stream, the program draws a square around the face. Now i want that much area of the image, in the square around the face, to be created as an another image. From what i've been doing, i'm getting a rectangular area of the image in which the face is not even present.

i've used the cv.GetSubRect() and understood its use. Like for eg:

img=cv.LoadImage("C:\opencv\me.jpg")
sub=cv.GetSubRect(img, (700,525,200,119))
cv.NamedWindow("result",1)
cv.ShowImage("result",sub)

• gives me the cropped picture of my eye.

But i can't get the face in my face and eye detection program. Here's what i've done:

min_size = (17,17)
        #max_size = (30,30)

        image_scale = 2
        haar_scale = 2

        min_neighbors = 2
        haar_flags = 0

        # Allocate the temporary images
        gray = cv.CreateImage((image.width, image.height), 8, 1)
        smallImage = cv.CreateImage((cv.Round(image.width / image_scale),cv.Round (image.height / image_scale)), 8 ,1)
        #eyeregion = cv.CreateImage((cv.Round(image.width / image_scale),cv.Round (image.height / image_scale)), 8 ,1)
        #cv.ShowImage("smallImage",smallImage)

        # Convert color input image to grayscale
        cv.CvtColor(image, gray, cv.CV_BGR2GRAY)

        # Scale input image for faster processing
        cv.Resize(gray, smallImage, cv.CV_INTER_LINEAR)

        # Equalize the histogram
        cv.EqualizeHist(smallImage, smallImage)

        # Detect the faces
        faces = cv.HaarDetectObjects(smallImage, faceCascade, cv.CreateMemStorage(0),
        haar_scale, min_neighbors, haar_flags, min_size)
        #, max_size)

# If faces are found
        if faces:
                for ((x, y, w, h), n) in faces:
                # the input to cv.HaarDetectObjects was resized, so scale the
                # bounding box of each face and convert it to two CvPoints
                        pt1 = (int(x * image_scale), int(y * image_scale))
                        pt2 = (int((x + w) * image_scale), int((y + h) * image_scale))
                        cv.Rectangle(image, pt1, pt2, cv.RGB(255, 0, 0), 3, 4, 0)
                        face_region = cv.GetSubRect(image,(x,int(y + (h/4)),w,int(h/2)))
                        cv.ShowImage("face",face_region)


 cv.SetImageROI(image, (pt1[0],
                        pt1[1],
                        pt2[0] - pt1[0],
                        int((pt2[1] - pt1[1]) * 0.7)))
                eyes = cv.HaarDetectObjects(image, eyeCascade,
                cv.CreateMemStorage(0),
                eyes_haar_scale, eyes_min_neighbors,
                eyes_haar_flags, eyes_min_size)    

            if eyes:
                    # For each eye found
                    for eye in eyes:

                            eye[0][0],eye[0][1] are x,y co-ordinates of the top-left corner of detected eye
                            eye[0][2],eye[0][3] are the width and height of the cvRect of the detected eye region (i mean c'mon, that can be made out from the for loop of the face detection)
                            # Draw a rectangle around the eye


                            ept1 = (eye[0][0],eye[0][1])
                            ept2 = ((eye[0][0]+eye[0][2]),(eye[0][1]+eye[0][3]))

                            cv.Rectangle(image,ept1,ept2,cv.RGB(0,0,255),1,8,0) # This is working.. 


                            ea = ept1[0]
                            eb = ept1[1]
                            ec = (ept2[0]-ept1[0])
                            ed = (ept2[1]-ept1[1])

                            # i've tried multiplying with image_scale to get the eye region within
                            # the window of eye but still i'm getting just a top-left area of the 
                            # image, top-left to my head. It does make sense to multiply with image_scale right?




                            eyeregion=cv.GetSubRect(image, (ea,eb,ec,ed))
                            cv.ShowImage("eye",eyeregion)

Answer 1

I hope this code is from OpenCV/samples/Python. There is a small mistake in the arguments you have given for the co-ordinates inside cv.GetSubRect. Please replace last two lines of above program with following:

a=pt1[0]
b=pt1[1]
c=pt2[0]-pt1[0]
d=pt2[1]-pt1[1]
face_region = cv.GetSubRect(image,(a,b,c,d))
cv.ShowImage("face",face_region) 

Make sure, you have no false detection or multiple detection.