def find_mask(img):
img.save(FingerprintIELib.FILE)
image = ndimage.imread(FingerprintIELib.FILE, mode="L").astype("float64")
image = Utils.normalize(image)
mask = Utils.find_mask(image)
misc.imsave(FingerprintIELib.FILE, mask)
return Image.open(FingerprintIELib.FILE)
def binarizing(img, filter_type, subdivide):
img.save(FingerprintIELib.FILE)
image = ndimage.imread(FingerprintIELib.FILE, mode="L").astype("float64")
image = Utils.normalize(image)
mask = Utils.find_mask(image)
image = np.where(mask == 1.0, Utils.local_normalize(image), image)
orientations = np.where(mask == 1.0, Utils.estimate_orientations(image), -1.0)
frequencies = np.where(mask == 1.0, Utils.estimate_frequencies(image, orientations), -1.0)
if filter_type == 'gabor':
if subdivide:
image = Utils.normalize(GaborUtil.gaborFilterSubdivide(image, orientations, frequencies))
else:
image = GaborUtil.gaborFilter(image, orientations, frequencies)
image = np.where(mask == 1.0, image, 1.0)
image = np.where(mask == 1.0, Utils.binarize(image), 1.0)
elif filter_type == 'wahab':
image = np.where(mask == 1.0, Utils.binarize(image), 1.0)
else:
raise SyntaxError('filter type %s is not supported' % filter_type)
misc.imsave(FingerprintIELib.FILE, image)
return Image.open(FingerprintIELib.FILE)
def gaborFilter(image, orientations, frequencies, w=32):
result = np.empty(image.shape)
height, width = image.shape
for y in range(0, height - w, w):
for x in range(0, width - w, w):
orientation = orientations[y+w//2, x+w//2]
frequency = Utils.average_frequency(frequencies[y:y + w, x:x + w])
if frequency < 0.0:
result[y:y+w, x:x+w] = image[y:y+w, x:x+w]
continue
kernel = GaborUtil.gaborKernel(16, orientation, frequency)
result[y:y+w, x:x+w] = Utils.convolve(image, kernel, (y, x), (w, w))
return Utils.normalize(result)
def gaborFilterSubdivide(image, orientations, frequencies, rect=None):
if rect:
y, x, h, w = rect
else:
y, x = 0, 0
h, w = image.shape
result = np.empty((h, w))
orientation, deviation = Utils.average_orientation(
orientations[y:y+h, x:x+w], deviation=True)
if (deviation < 0.2 and h < 50 and w < 50) or h < 6 or w < 6:
frequency = Utils.average_frequency(frequencies[y:y + h, x:x + w])
if frequency < 0.0:
result = image[y:y+h, x:x+w]
else:
kernel = GaborUtil.gaborKernel(16, orientation, frequency)
result = Utils.convolve(image, kernel, (y, x), (h, w))
else:
if h > w:
hh = h // 2
result[0:hh, 0:w] = \
GaborUtil.gaborFilterSubdivide(image, orientations, frequencies, (y, x, hh, w))
result[hh:h, 0:w] = \
GaborUtil.gaborFilterSubdivide(image, orientations, frequencies, (y + hh, x, h - hh, w))
else:
hw = w // 2
result[0:h, 0:hw] = \
GaborUtil.gaborFilterSubdivide(image, orientations, frequencies, (y, x, h, hw))
result[0:h, hw:w] = \
GaborUtil.gaborFilterSubdivide(image, orientations, frequencies, (y, x + hw, h, w - hw))
if w > 20 and h > 20:
result = Utils.normalize(result)
return result
def wahab_filter(image, orientations, w=8):
result = np.empty(image.shape)
height, width = image.shape
for y in range(0, height - w, w):
for x in range(0, width - w, w):
orientation = orientations[y + w // 2, x + w // 2]
kernel = WahabUtil.wahab_kernel(16, orientation)
result[y:y + w,
x:x + w] = Utils.convolve(image, kernel, (y, x), (w, w))
result[y:y + w, x:x + w] /= np.sum(kernel)
return result
def orientations(img):
img.save(FingerprintIELib.FILE)
image = ndimage.imread(FingerprintIELib.FILE, mode="L").astype("float64")
image = Utils.normalize(image)
mask = Utils.find_mask(image)
image = np.where(mask == 1.0, Utils.local_normalize(image), image)
orientations = np.where(mask == 1.0, Utils.estimate_orientations(image), -1.0)
Utils.show_orientations(image, orientations, "orientations", 8)
plt.savefig(FingerprintIELib.FILE, bbox_inches='tight', pad_inches = 0)
return Image.open(FingerprintIELib.FILE)
def gaborKernel(size, angle, frequency):
"""
Create a Gabor kernel given a size, angle and frequency.
Code is taken from https://github.com/rtshadow/biometrics.git
"""
angle += np.pi * 0.5
cos = np.cos(angle)
sin = -np.sin(angle)
yangle = lambda x, y: x * cos + y * sin
xangle = lambda x, y: -x * sin + y * cos
xsigma = ysigma = 4
return Utils.kernel_from_function(size, lambda x, y:
np.exp(-(
(xangle(x, y) ** 2) / (xsigma ** 2) +
(yangle(x, y) ** 2) / (ysigma ** 2)) / 2) *
np.cos(2 * np.pi * frequency * xangle(x, y)))