def feature_stats_test():
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
print(np.mean(X_data, axis=0))
print(np.std(X_data, axis=0))
def extract_coordinate_feature_test():
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
fig = plt.figure(figsize=(10,10))
ax1 = fig.add_subplot(121)
ax1.imshow(I)
ax2 = fig.add_subplot(122)
ax2.imshow(coord_im)
def extract_features(image_number, slice_number):
# extracts features for [image_number]_[slice_number]_t1.tif and [image_number]_[slice_number]_t2.tif
# Input:
# image_number - Which subject (scalar)
# slice_number - Which slice (scalar)
# Output:
# X - N x k dataset, where N is the number of pixels and k is the total number of features
# features - k x 1 cell array describing each of the k features
base_dir = '../data/dataset_brains/'
t1 = plt.imread(base_dir + str(image_number) + '_' + str(slice_number) +
'_t1.tif')
t1g = scipy.ndimage.gaussian_filter(t1, sigma=1)
s1x = scipy.ndimage.sobel(t1, axis=0, mode='constant')
s1y = scipy.ndimage.sobel(t1, axis=1, mode='constant')
t1s = np.hypot(s1x, s1y)
t2 = plt.imread(base_dir + str(image_number) + '_' + str(slice_number) +
'_t2.tif')
t2g = scipy.ndimage.gaussian_filter(t2, sigma=1)
s2x = scipy.ndimage.sobel(t2, axis=0, mode='constant')
s2y = scipy.ndimage.sobel(t2, axis=1, mode='constant')
t2s = np.hypot(s2x, s2y)
n = t1.shape[0]
features = ()
t1f = t1.flatten().T.astype(float).reshape(-1, 1)
t1gf = t1g.flatten().T.astype(float).reshape(-1, 1)
t1sf = t1s.flatten().T.astype(float).reshape(-1, 1)
t2f = t2.flatten().T.astype(float).reshape(-1, 1)
t2gf = t2g.flatten().T.astype(float).reshape(-1, 1)
t2sf = t2s.flatten().T.astype(float).reshape(-1, 1)
t_diff = np.abs(t1f - t2f)
r, _ = seg.extract_coordinate_feature(t1)
X = np.concatenate((t1f, t2f), axis=1)
X = np.concatenate((X, t1gf), axis=1)
X = np.concatenate((X, t2gf), axis=1)
X = np.concatenate((X, t1sf), axis=1)
X = np.concatenate((X, t2sf), axis=1)
X = np.concatenate((X, t_diff), axis=1)
X = np.concatenate((X, r), axis=1)
features += ('T1 intensity', )
features += ('T2 intensity', )
features += ('T1 intensity gaussian filter', )
features += ('T2 intensity gaussian filter', )
features += ('T1 intensity sobel filter', )
features += ('T2 intensity sobel filter', )
features += ('abs(T1 - T2)', )
features += ('distance to center', )
return X, features
def feature_stats_test():
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
print("=========== Feature Stats Test ===========\n")
for i in range(np.size(X_data[0,:])):
mean = np.mean(X_data[:, i])
std = np.std(X_data[:, i])
print("Feature ", i+1, "\nMean: \t", round(mean, 3), "\nStd: \t", round(std, 3), "\n")
print("=================== END ==================")
def kmeans_clustering_test():
#------------------------------------------------------------------#
#TODO: Store errors for training data
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
test_data, _ = seg.normalize_data(X_data)
predicted_labels = seg.kmeans_clustering(test_data)
predicted_labels = predicted_labels.reshape(I.shape)
plt.imshow(predicted_labels)
def feature_stats_test():
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
#------------------------------------------------------------------#
# TODO: Write code to examine the mean and standard deviation of your dataset containing variety of features
mean = np.mean(X_data, axis=0) #berekent mean per rij
standard_deviation = np.std(X_data, axis=0)
print("Mean is" + str(mean))
print("Std is" + str(standard_deviation))
def kmeans_clustering_test():
#------------------------------------------------------------------#
#TODO: Store errors for training data
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
#
normalized_Xdata, _ = seg.normalize_data(X_data)
kmeans_cost = seg.kmeans_clustering(normalized_Xdata)
return kmeans_cost
def feature_stats_test():
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
#------------------------------------------------------------------#
# TODO: Write code to examine the mean and standard deviation of your dataset containing variety of features
feature_mean = np.mean(X_data,0)
feature_std = np.std(X_data,0)
for i in range(6):
print("Feature {} has the following properties. The mean is: {:.2f} and the standard deviation is: {:.2f}".format(i+1,feature_mean.item(i), feature_std.item(i)))
def feature_stats_test():
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
#------------------------------------------------------------------#
# Write code to examine the mean and standard deviation of your dataset containing variety of features
means = np.zeros(X_data.shape[1])
stds = np.zeros(X_data.shape[1])
for i in range(X_data.shape[1]):
means[i] = np.mean(X_data[:, i])
stds[i] = np.std(X_data[:, i])
util.scatter_data(X_data, Y, 0, 1)
print(means)
print(stds)
def normalized_stats_test():
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
#------------------------------------------------------------------#
# TODO: Write code to normalize your dataset containing variety of features,
# then examine the mean and std dev
normdata, _ = seg.normalize_data(
X_data) #output van def normalize_data is 2 variables
mean = np.mean(normdata, axis=0)
standard_deviation = np.std(normdata, axis=0)
print("Mean is" + str(mean))
print("Std is" + str(standard_deviation))
def normalized_stats_test():
X, Y = scatter_data_test(showFigs=False)
I = plt.imread('../data/dataset_brains/1_1_t1.tif')
c, coord_im = seg.extract_coordinate_feature(I)
X_data = np.concatenate((X, c), axis=1)
def extract_features(image_number, slice_number):
# extracts features for [image_number]_[slice_number]_t1.tif and [image_number]_[slice_number]_t2.tif
# Input:
# image_number - Which subject (scalar)
# slice_number - Which slice (scalar)
# Output:
# X - N x k dataset, where N is the number of pixels and k is the total number of features
# features - k x 1 cell array describing each of the k features
base_dir = '../data/dataset_brains/'
t1 = plt.imread(base_dir + str(image_number) + '_' + str(slice_number) +
'_t1.tif')
t2 = plt.imread(base_dir + str(image_number) + '_' + str(slice_number) +
'_t2.tif')
n = t1.shape[0]
features = ()
t1f = t1.flatten().T.astype(float)
t1f = t1f.reshape(-1, 1)
t2f = t2.flatten().T.astype(float)
t2f = t2f.reshape(-1, 1)
X = np.concatenate((t1f, t2f), axis=1)
features += ('T1 intensity', )
features += ('T2 intensity', )
#------------------------------------------------------------------#
# Extract more features and add them to X.
# Don't forget to provide (short) descriptions for the features
# Feature 3: T1 Gaussian filtered
I_blurred = ndimage.gaussian_filter(t1, sigma=10)
X2 = I_blurred.flatten().T
t1_blur_10 = X2.reshape(-1, 1)
features += ('T1 Intensity (Gaussian)', )
# Feature 4: T2 Gaussian filtered
I_blurred_t2 = ndimage.gaussian_filter(t2, sigma=10)
X3 = I_blurred_t2.flatten().T
t2_blur_10 = X3.reshape(-1, 1)
features += ('T2 Intensity (Gaussian)', )
# Feature 5: T1 Median filtered
t1_med = pro.create_my_feature(t1)
t1f_med = t1_med.flatten().T
t1f_med = t1f_med.reshape(-1, 1)
features += ('T1 Intensity (Median)', )
# Feature 6: T2 Median filtered
t2_med = pro.create_my_feature(t2)
t2f_med = t2_med.flatten().T
t2f_med = t2f_med.reshape(-1, 1)
features += ('T2 Intensity (Median)', )
# Feature 7: T1-T2 absolute difference
t_diff = np.sqrt((t1 - t2) ^ 2)
tf_diff = t_diff.flatten().T.astype(float)
tf_diff = tf_diff.reshape(-1, 1)
features += ('T1-T2 difference', )
# Feature 8: Distance to center
c, c_im = seg.extract_coordinate_feature(t1)
features += ('Distance to Center', )
X = np.concatenate(
(X, t1_blur_10, t2_blur_10, t1f_med, t2f_med, tf_diff, c), axis=1)
#------------------------------------------------------------------#
return X, features
t1_blurred_1 = ndimage.gaussian_filter(t1, sigma=3)
t1_1 = t1_blurred_1.flatten().T
t1_1 = t1_1.reshape(-1, 1)
t1_blurred_2 = ndimage.gaussian_filter(t1, sigma=8)
t1_2 = t1_blurred_2.flatten().T
t1_2 = t1_2.reshape(-1, 1)
#edges
t1_LaPlacian = cv2.Laplacian(t1, cv2.CV_64F)
t1_lapl = t1_LaPlacian.flatten().T
t1_lapl = t1_lapl.reshape(-1, 1)
#Coordinate feature
t1_coord, _ = seg.extract_coordinate_feature(t1)
#use to remove noise from images, while preserving edges
median = cv2.medianBlur(t1, 5)
t1_median = median.flatten().flatten().T
t1_median = t1_median.reshape(-1, 1)
#opening
kernel1 = np.array([[0, 1, 0], [1, 1, 1], [0, 1, 0]],
np.uint8) #np.ones((3,3), np.uint8)
opening = cv2.morphologyEx(t1, cv2.MORPH_OPEN, kernel1)
t1_opening = opening.flatten().flatten().T
t1_opening = t1_opening.reshape(-1, 1)
#closing
kernel2 = np.ones((4, 4), np.uint8)
def extract_myfeatures(image_number, slice_number):
# extracts features for [image_number]_[slice_number]_t1.tif and [image_number]_[slice_number]_t2.tif
# Input:
# image_number - Which subject (scalar)
# slice_number - Which slice (scalar)
# Output:
# X - N x k dataset, where N is the number of pixels and k is the total number of features
# features - k x 1 cell array describing each of the k features
base_dir = '../data/dataset_brains/'
t1 = plt.imread(base_dir + str(image_number) + '_' + str(slice_number) +
'_t1.tif')
n = t1.shape[0]
features = ()
t1f = t1.flatten().T.astype(float)
t1f = t1f.reshape(-1, 1)
features += ('T1 intensity', )
#------------------------------------------------------------------#
# TODO: Extract more features and add them to X.
# Don't forget to provide (short) descriptions for the features
#Features for T1
t1_blurred_1 = ndimage.gaussian_filter(t1, sigma=3)
t1_1 = t1_blurred_1.flatten().T
t1_1 = t1_1.reshape(-1, 1)
t1_blurred_2 = ndimage.gaussian_filter(t1, sigma=8)
t1_2 = t1_blurred_2.flatten().T
t1_2 = t1_2.reshape(-1, 1)
#edges
t1_LaPlacian = cv2.Laplacian(t1, cv2.CV_64F)
t1_lapl = t1_LaPlacian.flatten().T
t1_lapl = t1_lapl.reshape(-1, 1)
#Coordinate feature
t1_coord, _ = seg.extract_coordinate_feature(t1)
#use to remove noise from images, while preserving edges
median = cv2.medianBlur(t1, 5)
t1_median = median.flatten().flatten().T
t1_median = t1_median.reshape(-1, 1)
#opening
kernel1 = np.array([[0, 1, 0], [1, 1, 1], [0, 1, 0]],
np.uint8) #np.ones((3,3), np.uint8)
opening = cv2.morphologyEx(t1, cv2.MORPH_OPEN, kernel1)
t1_opening = opening.flatten().flatten().T
t1_opening = t1_opening.reshape(-1, 1)
#closing
kernel2 = np.ones((4, 4), np.uint8)
closing = cv2.morphologyEx(t1, cv2.MORPH_CLOSE, kernel2)
t1_closing = closing.flatten().flatten().T
t1_closing = closing.reshape(-1, 1)
#morphological gradient
kernel3 = np.ones((3, 3), np.uint8)
gradient = cv2.morphologyEx(t1, cv2.MORPH_GRADIENT, kernel3)
t1_grad = gradient.flatten().flatten().T
t1_grad = t1_grad.reshape(-1, 1)
t1_cv = cv2.imread(
base_dir + str(image_number) + '_' + str(slice_number) + '_t1.tif', 0)
equ = cv2.equalizeHist(t1_cv)
cv2.imwrite('equ.png', equ)
t1_equ = plt.imread('equ.png')
t1_equ = t1_equ.flatten().T
t1_equ = t1_equ.reshape(-1, 1)
#Add features to the list of features
features += ('T1 gauss 5', )
features += ("T1 equalized histogram", )
features += ('T1 gauss 15', )
features += ("T1 Coordinate feature", )
features += ("T1 median", )
features += ("T1 opening", )
features += ("T1 closing", )
features += ("T1 morphological gradient", )
features += ("T1 Laplacian", )
X = np.concatenate((t1f, t1_equ, t1_1, t1_1, t1_coord, t1_median,
t1_opening, t1_closing, t1_grad, t1_lapl),
axis=1)
#------------------------------------------------------------------#
return X, features
