def __init__(self,filename,color,cascade,READ_FILE):
labeled=zip(filename,color)
label=[]
feature_train=[]
for (fd,c) in labeled:
print fd
if READ_FILE is True:
img=cv2.imread(fd,0)
else:
img = filename
if img is None:
print "image can't be loaded in face recog"
sys.exit(0)
face_cascade=cv2.CascadeClassifier(cascade)
features=Feature_extractor(img)
face=face_cascade.detectMultiScale(
img,
scaleFactor=1.1,
minNeighbors=2,
minSize=(10,10)
)
window=face[0]
(x,y,w,h)=window
#print "selectkp"
#print window
kp=features.keypoints()
selectKp=features.select_kp(window)
#print selectKp
for (X,Y) in selectKp:
plt.scatter(X,Y,color=c)
label.append(c)
feature_train.append([X,Y])
#cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,0),1)
#cv2.imshow("image",img)
self.label=label
self.feature_train=feature_train
__author__ = 'Abdullah_Rahman'
#r=kath
threshold=0.53
filenames=("aron.jpg","sharapova_test.jpg")
color=('r','b')
cascade="cascades/frontalface.xml"
C=KMeans(n_clusters=4)
face_cascade=cv2.CascadeClassifier(cascade)
image=cv2.imread("sharapova_test.jpg",0)
features=Feature_extractor(image)
if image is None:
print "IMAGE NOT FOUND in input_image.py"
sys.exit(0)
face=face_cascade.detectMultiScale(
image,
scaleFactor=1.1,
minNeighbors=2,
minSize=(400,400)
)
window=face[0]
return zip([i for i in range(len(A))],A)
print "READING THE IMAGE FILE AND EXTRACTING KEYPOINTS"
input = cv2.imread("aron.jpg",0)
height = np.size(input,0)
width = np.size(input,1)
print height,width
cascade = cv2.CascadeClassifier('cascades/frontalface.xml')
model_filename = "TRAINED_SUPPORT_VECTOR_MACHINE.pkl"
model = pickle.load(open(model_filename,'rb'))
features = Feature_extractor(input)
pt = features.keypoints()
normalized=[(100*(x)/width,100*(y)/height) for (x,y) in pt]
pt=[(int(x),int(y)) for (x,y) in normalized]
print len(pt)
print "ANALYZING FEATURES"
Kp = np.asarray(pt)
segments = separate(Kp,50,50)
ratios = ratio_face(segments)
c ={1:'r',0:'b'}
X1=np.linspace(0,45,40)
Y1=np.linspace(0.5,1.5,40)
for x in X1: