def take_snapshot(img, iteration_step=0):
"""
Given image and iteration_step, saves 'image' concatenated with
'iteration_step'.
This function is ONLY for debug purposes.
"""
folder_name = "./Output/"
filename = "image_isodata" + "_" + str(iteration_step)
img_eq = equalization_using_histogram(img)
save_image(folder_name, filename, img_eq)
def take_snapshot(img, iteration_step=0):
"""
Given image and iteration_step, saves 'image' concatenated with
'iteration_step'.
This function is ONLY for debug purposes.
"""
folder_name = "./Output/"
filename = "image_isodata" + "_" + str(iteration_step)
img_eq = equalization_using_histogram(img)
save_image(folder_name, filename, img_eq)
def iso(image):
# this should be placed at the top with all the imports
from pyradar.classifiers.isodata import isodata_classification
from pyradar.core.equalizers import equalization_using_histogram
from pyradar.core.sar import save_image
params = {"K": 15, "I": 100, "P": 2, "THETA_M": 10, "THETA_S": 0.1,
"THETA_C": 2, "THETA_O": 0.01}
# run Isodata
class_image = isodata_classification(image, parameters=params)
# equalize class image to 0:255
class_image_eq = equalization_using_histogram(class_image)
show_image(class_image_eq, 'isodata_classification')
def kmeans(image):
from pyradar.classifiers.kmeans import kmeans_classification
# number of clusters
k = 4
# max number of iterations
iter_max = 1000
# run K-Means
class_image = kmeans_classification(image, k, iter_max)
# equalize class image to 0:255
class_image_eq = equalization_using_histogram(class_image)
# save it
show_image(class_image_eq, 'isodata_classification')
def kmeans(image):
from pyradar.classifiers.kmeans import kmeans_classification
# number of clusters
k = 4
# max number of iterations
iter_max = 1000
# run K-Means
class_image = kmeans_classification(image, k, iter_max)
# equalize class image to 0:255
class_image_eq = equalization_using_histogram(class_image)
# save it
show_image(class_image_eq, 'isodata_classification')
def iso(image):
# this should be placed at the top with all the imports
from pyradar.classifiers.isodata import isodata_classification
from pyradar.core.equalizers import equalization_using_histogram
from pyradar.core.sar import save_image
params = {
"K": 15,
"I": 100,
"P": 2,
"THETA_M": 10,
"THETA_S": 0.1,
"THETA_C": 2,
"THETA_O": 0.01
}
# run Isodata
class_image = isodata_classification(image, parameters=params)
# equalize class image to 0:255
class_image_eq = equalization_using_histogram(class_image)
show_image(class_image_eq, 'isodata_classification')
def eq(image):
# equalize img to 0:255
image_eq = equalization_using_histogram(image)
# save img in current directory
show_image(image_eq, 'Equalized image via Hist')
def export_image_layer(self, layer_name, filename, path_to):
""" Export an image file for a given layer to the given path. """
layer = self._get_layer_securely(layer_name)
img_matrix = equalization_using_histogram(layer)
save_image(path_to, filename, img_matrix)
def test_all(self):
image_path = os.path.join(IMAGES, "DAT_01.001")
print image_path
dataset = create_dataset_from_path(image_path)
band = get_band_from_dataset(dataset)
geoinfo = get_geoinfo(dataset, cast_to_int=True)
xoff = geoinfo['xoff'] + 2000
yoff = geoinfo['yoff'] + 2000
## Parameters:
win_xsize = 100 # window size in coord x
win_ysize = 100 # window size in coord y
k = 1 # parameter of frost filter, ex: k=1 or k=10 or k=100
win_size = 3 # size of the window for the filter function
damping_factor = 1 # parameter of frost filter, ex: 1 or 10 or 1000
image = read_image_from_band(band, xoff, yoff, win_xsize, win_ysize)
# Try K-Means
kmean_timer = Timer()
n_classes = 8
iterations = 1000
class_image = kmeans_classification(image, n_classes, iterations)
kmean_timer.stop_timer()
kmean_timer.calculate_time_elapsed(print_value=True)
# Try Isodata
isodata_timer = Timer()
parameters={"K": 8, "I":1000}
class_image = isodata_classification(image,parameters=parameters )
isodata_timer.stop_timer()
isodata_timer.calculate_time_elapsed(print_value=True)
numerito = parameters["K"]
# Try the filters
filter_timer = Timer()
numerito = 11
cucito = 0.30
image_filtered = mean_filter(image, win_size=numerito)
image_filtered = median_filter(image,win_size)
image_filtered = frost_filter(image, damping_factor=1.0, win_size=11)
image_filtered = kuan_filter(image, win_size=7, cu=1.0)
image_filtered = lee_filter(image, win_size=numerito, cu=cucito)
image_filtered = lee_enhanced_filter(image, win_size=numerito, cu=cucito)
filter_timer.stop_timer()
diff = filter_timer.calculate_time_elapsed(print_value=True)
# Frost y K-Means
ventana = 7
damp = 1.0
#
parameters={"K" : 5, "I" : 100}
#
_timer = Timer()
#
image_filtered = frost_filter(image, damping_factor=damp, win_size=ventana)
class_image = isodata_classification(image_filtered, parameters)
#
#
_timer.stop_timer()
_timer.calculate_time_elapsed(print_value=True)
#
image_corrected = equalization_using_histogram(class_image)
save_image(IMG_DEST_DIR, "image_" + str(ventana) + "frostisodata" +
str(damp) + "c" + str(parameters["K"]), image_corrected)
#
#
# Equalize and save the images to files
image_corrected = equalization_using_histogram(class_image)
save_image(IMG_DEST_DIR, "image_isodata8", image_corrected)
#
image_original = equalization_using_histogram(image)
save_image(IMG_DEST_DIR, "image_original", image_original)
print "\a\a\a"
def export_image_layer(self, layer_name, filename, path_to):
""" Export an image file for a given layer to the given path. """
layer = self._get_layer_securely(layer_name)
img_matrix = equalization_using_histogram(layer)
save_image(path_to, filename, img_matrix)
def eq(image):
# equalize img to 0:255
image_eq = equalization_using_histogram(image)
# save img in current directory
show_image(image_eq, 'Equalized image via Hist')
def test_all(self):
image_path = os.path.join(IMAGES, "DAT_01.001")
print image_path
dataset = create_dataset_from_path(image_path)
band = get_band_from_dataset(dataset)
geoinfo = get_geoinfo(dataset, cast_to_int=True)
xoff = geoinfo['xoff'] + 2000
yoff = geoinfo['yoff'] + 2000
## Parameters:
win_xsize = 100 # window size in coord x
win_ysize = 100 # window size in coord y
k = 1 # parameter of frost filter, ex: k=1 or k=10 or k=100
win_size = 3 # size of the window for the filter function
damping_factor = 1 # parameter of frost filter, ex: 1 or 10 or 1000
image = read_image_from_band(band, xoff, yoff, win_xsize, win_ysize)
# Try K-Means
kmean_timer = Timer()
n_classes = 8
iterations = 1000
class_image = kmeans_classification(image, n_classes, iterations)
kmean_timer.stop_timer()
kmean_timer.calculate_time_elapsed(print_value=True)
# Try Isodata
isodata_timer = Timer()
parameters={"K": 8, "I":1000}
class_image = isodata_classification(image,parameters=parameters )
isodata_timer.stop_timer()
isodata_timer.calculate_time_elapsed(print_value=True)
numerito = parameters["K"]
# Try the filters
filter_timer = Timer()
numerito = 11
cucito = 0.30
image_filtered = mean_filter(image, win_size=numerito)
image_filtered = median_filter(image,win_size)
image_filtered = frost_filter(image, damping_factor=1.0, win_size=11)
image_filtered = kuan_filter(image, win_size=7, cu=1.0)
image_filtered = lee_filter(image, win_size=numerito, cu=cucito)
image_filtered = lee_enhanced_filter(image, win_size=numerito, cu=cucito)
filter_timer.stop_timer()
diff = filter_timer.calculate_time_elapsed(print_value=True)
# Frost y K-Means
ventana = 7
damp = 1.0
#
parameters={"K" : 5, "I" : 100}
#
_timer = Timer()
#
image_filtered = frost_filter(image, damping_factor=damp, win_size=ventana)
class_image = isodata_classification(image_filtered, parameters)
#
#
_timer.stop_timer()
_timer.calculate_time_elapsed(print_value=True)
#
image_corrected = equalization_using_histogram(class_image)
save_image(IMG_DEST_DIR, "image_" + str(ventana) + "frostisodata" +
str(damp) + "c" + str(parameters["K"]), image_corrected)
#
#
# Equalize and save the images to files
image_corrected = equalization_using_histogram(class_image)
save_image(IMG_DEST_DIR, "image_isodata8", image_corrected)
#
image_original = equalization_using_histogram(image)
save_image(IMG_DEST_DIR, "image_original", image_original)
print "\a\a\a"
from pyradar.core.sar import read_image_from_band from pyradar.core.sar import save_image from pyradar.core.equalizers import equalization_using_histogram IMAGE_PATH = "./img_sar/DAT_01.001" IMG_DEST_DIR = "." # create dataset dataset = create_dataset_from_path(IMAGE_PATH) # get band from dataset band = get_band_from_dataset(dataset) # get geo info from dataset geoinfo = get_geoinfo(dataset, cast_to_int=True) #usually both values are zero xoff = geoinfo['xoff'] yoff = geoinfo['yoff'] # window size in coord x win_xsize = 128 # window size in coord y win_ysize = 128 image = read_image_from_band(band, xoff, yoff, win_xsize, win_ysize) #equalize img to 0:255 image_eq = equalization_using_histogram(image) # save img in current directory save_image(IMG_DEST_DIR, "image_sar", image_eq)
from pyradar.core.sar import save_image from pyradar.core.equalizers import equalization_using_histogram IMAGE_PATH = "./img_sar/DAT_01.001" IMG_DEST_DIR = "." # create dataset dataset = create_dataset_from_path(IMAGE_PATH) # get band from dataset band = get_band_from_dataset(dataset) # get geo info from dataset geoinfo = get_geoinfo(dataset, cast_to_int=True) #usually both values are zero xoff = geoinfo['xoff'] yoff = geoinfo['yoff'] # window size in coord x win_xsize = 128 # window size in coord y win_ysize = 128 image = read_image_from_band(band, xoff, yoff, win_xsize, win_ysize) #equalize img to 0:255 image_eq = equalization_using_histogram(image) # save img in current directory save_image(IMG_DEST_DIR, "image_sar", image_eq)
params = {
"K": 10,
"I": 100,
"P": 0,
"THETA_M": 1,
"THETA_S": 0.1,
"THETA_C": 2,
"THETA_O": 0.01
}
IMAGE_PATH = "/home/rccc/Downloads/LC81360272014242LGN00_B1.TIF"
IMG_DEST_DIR = "/home/rccc/Downloads/"
# create dataset
#dataset = create_dataset_from_path(IMAGE_PATH)
dataset = misc.imread(IMAGE_PATH)
# run Isodata
class_image = isodata_classification(dataset, parameters=params)
# equalize class image to 0:255
class_image_eq = equalization_using_histogram(class_image)
# save it
save_image(IMG_DEST_DIR, "class_image_eq", class_image_eq)
# also save original image
image_eq = equalization_using_histogram(dataset)
# save it
save_image(IMG_DEST_DIR, "image_eq", image_eq)
