def __init__(self,
images_dir,
picture_amount_threshold,
feature_method_list,
row_threshold,
column_threshold,
picture_resolution,
label_threshold,
ground_truth_dir,
abs_flag,
log_flag,
grey_picture=False,
auto_canny=True,
auto_canny_sigma=0.33):
saved_args = locals()
Arguments.logArguments('FeatureDataSet', saved_args)
self.create_grey_dir = False
self.images_dir = images_dir
self.picture_amount_threshold = picture_amount_threshold # number of pictures
self.grey_picture = grey_picture # load picture as grey/color
self.feature_method_list = feature_method_list # list of feature extraction method
self.row_threshold = row_threshold
self.column_threshold = column_threshold
self.picture_resolution = picture_resolution
self.label_threshold = label_threshold # min amount of label 1 percentage
self.ground_truth_dir = ground_truth_dir # dir of ground true pictures - extract only high class 1 percent
self.abs_flag = abs_flag
self.log_flag = log_flag
self.images_pixel_dict = dict() # contain images and their pixels
self.images_name_list = list() # contain pic name in data set
self.X = list() # list of ndarray - feature format
self.X_sobelx = list()
self.X_sobely = list()
self.X_laplacian = list()
self.X_canny = list()
self.X_scharrx = list()
self.X_scharry = list()
self.X_sobelx_final = list()
self.X_sobely_final = list()
self.X_laplacian_final = list()
self.X_canny_final = list()
self.X_scharrx_final = list()
self.X_scharry_final = list()
self.auto_canny = auto_canny
self.auto_canny_sigma = auto_canny_sigma
if self.auto_canny_sigma < 0 or self.auto_canny_sigma > 1:
raise 'Illegal value for auto Canny Sigma parameter - need to be 0<=..<=1'
self.kepsilon = 1e-8
return
def __init__(self,
crf,
w,
image_name_list,
X,
ground_truth_dir,
pixel_frame,
evaluation_visualisation_flag,
dir_visualization_name,
model_name,
feature_data_set_obj,
baseline_features=None):
saved_args = locals()
del saved_args['X'] # preventing from logging
Arguments.logArguments('Evaluate', saved_args)
self.crf = crf
self.w = w
self.X = X
self.image_name_list = image_name_list
self.ground_truth_dir = ground_truth_dir
self.pixels_frame = pixel_frame
self.evaluation_visualisation_flag = evaluation_visualisation_flag
self.dir_visualization_name = dir_visualization_name
self.model_name = model_name
self.baseline_features_dict = dict()
if len(baseline_features) > 0:
for f in baseline_features:
attr = getattr(feature_data_set_obj, 'X_{}_final'.format(f))
self.baseline_features_dict[f] = attr
self.all_f1_avg = list() # avg f1 (between all gt)
self.all_f1 = list() # max f1 (max per gt)
self.all_recall = list()
self.all_precision = list()
self.all_super_f1_avg = list()
self.all_super_f1 = list()
self.all_super_recall = list()
self.all_super_precision = list()
self.canny_f1_avg = list()
self.canny_f1 = list()
self.canny_recall = list()
self.canny_precision = list()
self.canny_super_f1_avg = list()
self.canny_super_f1 = list()
self.canny_super_recall = list()
self.canny_super_precision = list()
return
def __init__(self, ground_truth_dir, images_name_list, row_threshold, column_threshold, X, pixels_frame):
saved_args = locals()
del saved_args['X'] # preventing from logging
Arguments.logArguments('TargetDataSet', saved_args)
self.ground_truth_dir = ground_truth_dir
self.images_name_list = images_name_list
self.row_threshold = row_threshold
self.column_threshold = column_threshold
self.pixels_frame = pixels_frame
self.y = list()
self.X = X
self.delete_indexes = list()
return
def __init__(self, X, y, learners, learners_parameters, models, models_parameters, total_label_one_avg, row_threshold):
saved_args = locals()
del saved_args['X'] # preventing from logging
del saved_args['y'] # preventing from logging
Arguments.logArguments('Model', saved_args)
self.X = X
self.y = y
self.learners = learners
self.learners_parameters = learners_parameters
self.models = models
self.models_parameters = models_parameters
self.total_label_one_avg = total_label_one_avg
self.row_threshold = row_threshold # needed to define graph model
self.total_accuracy_list = list() # each image and relevance accuracy
return
