From c35479d3dc7f33e2de28fb7c55ac7ae4411f792b Mon Sep 17 00:00:00 2001
From: Andre Anjos <andre.anjos@idiap.ch>
Date: Tue, 26 Jul 2016 15:18:27 +0200
Subject: [PATCH] [preproc] Improve finger post-processing (heq); Fix warnings
---
bob/bio/vein/preprocessors/FingerCrop.py | 58 ++++++++++++++++--------
1 file changed, 39 insertions(+), 19 deletions(-)
diff --git a/bob/bio/vein/preprocessors/FingerCrop.py b/bob/bio/vein/preprocessors/FingerCrop.py
index 89baf20..2285afb 100644
--- a/bob/bio/vein/preprocessors/FingerCrop.py
+++ b/bob/bio/vein/preprocessors/FingerCrop.py
@@ -26,10 +26,10 @@ class FingerCrop (Preprocessor):
In this implementation, the finger image is (in this order):
- 1. Padded
- 2. The mask is extracted
- 3. The finger is normalized (made horizontal)
- 3. (optionally) Post processed
+ 1. Padded
+ 2. The mask is extracted
+ 3. The finger is normalized (made horizontal)
+ 4. (optionally) Post processed
Parameters:
@@ -220,15 +220,15 @@ class FingerCrop (Preprocessor):
# Right region has always the finger ending, crop the padding with the
# meadian
- finger_mask[:,numpy.median(y_rg)+img_filt_rg.shape[1]:] = False
+ finger_mask[:,int(numpy.median(y_rg)+img_filt_rg.shape[1]):] = False
# Extract y-position of finger edges
edges = numpy.zeros((2,img_w))
edges[0,:] = y_up
- edges[0,0:round(numpy.mean(y_lf))+1] = edges[0,round(numpy.mean(y_lf))+1]
+ edges[0,0:int(round(numpy.mean(y_lf))+1)] = edges[0,int(round(numpy.mean(y_lf))+1)]
edges[1,:] = numpy.round(y_lo + img_filt_lo.shape[0])
- edges[1,0:round(numpy.mean(y_lf))+1] = edges[1,round(numpy.mean(y_lf))+1]
+ edges[1,0:int(round(numpy.mean(y_lf))+1)] = edges[1,int(round(numpy.mean(y_lf))+1)]
return finger_mask, edges
@@ -401,8 +401,13 @@ class FingerCrop (Preprocessor):
return (image_norm, mask_norm)
- def __HE__(self, image):
- """Applies histogram equalization on the input image
+ def __HE__(self, image, mask):
+ """Applies histogram equalization on the input image inside the mask
+
+ In this implementation, only the pixels that lie inside the mask will be
+ used to calculate the histogram equalization parameters. Because of this
+ particularity, we don't use Bob's implementation for histogram equalization
+ and have one based exclusively on NumPy.
Parameters:
@@ -410,6 +415,10 @@ class FingerCrop (Preprocessor):
image (numpy.ndarray): raw image to be filtered, as 2D array of
unsigned 8-bit integers
+ mask (numpy.ndarray): mask of the same size of the image, but composed
+ of boolean values indicating which values should be considered for
+ the histogram equalization
+
Returns:
@@ -418,9 +427,19 @@ class FingerCrop (Preprocessor):
"""
- #Umbralization based on the pixels non zero
- retval = numpy.zeros(image.shape, dtype=numpy.uint8)
- bob.ip.base.histogram_equalization(image, retval)
+ image_histogram, bins = numpy.histogram(image[mask], 256, normed=True)
+ cdf = image_histogram.cumsum() # cumulative distribution function
+ cdf = 255 * cdf / cdf[-1] # normalize
+
+ # use linear interpolation of cdf to find new pixel values
+ image_equalized = numpy.interp(image.flatten(), bins[:-1], cdf)
+ image_equalized = image_equalized.reshape(image.shape)
+
+ # normalized image to be returned is a composition of the original image
+ # (background) and the equalized image (finger area)
+ retval = image.copy()
+ retval[mask] = image_equalized[mask]
+
return retval
@@ -521,32 +540,33 @@ class FingerCrop (Preprocessor):
"""Reads the input image, extract the mask of the fingervein, postprocesses
"""
+ import ipdb; ipdb.set_trace()
+
# 1. Pads the input image if any padding should be added
image = numpy.pad(image, self.padding_width, 'constant',
constant_values = self.padding_constant)
## Finger edges and contour extraction:
if self.fingercontour == 'leemaskMatlab':
- finger_mask, finger_edges = self.__leemaskMatlab__(image) #for UTFVP
+ mask, edges = self.__leemaskMatlab__(image) #for UTFVP
elif self.fingercontour == 'leemaskMod':
- finger_mask, finger_edges = self.__leemaskMod__(image) #for VERA
+ mask, edges = self.__leemaskMod__(image) #for VERA
elif self.fingercontour == 'konomask':
- finger_mask, finger_edges = self.__konomask__(image, sigma=5)
+ mask, edges = self.__konomask__(image, sigma=5)
## Finger region normalization:
- image_norm, finger_mask_norm = self.__huangnormalization__(image,
- finger_mask, finger_edges)
+ image_norm, mask_norm = self.__huangnormalization__(image, mask, edges)
## veins enhancement:
if self.postprocessing == 'HE':
- image_norm = self.__HE__(image_norm)
+ image_norm = self.__HE__(image_norm, mask_norm)
elif self.postprocessing == 'HFE':
image_norm = self.__HFE__(image_norm)
elif self.postprocessing == 'CircGabor':
image_norm = self.__circularGabor__(image_norm, 1.12, 5)
## returns the normalized image and the finger mask
- return image_norm, finger_mask_norm
+ return image_norm, mask_norm
def write_data(self, data, filename):
--
GitLab