from fcn_model import createModel
from inputParser import parse_input
from normalization import normalize_scans
from metrics import dice
import numpy as np
import time
import keras
from keras.preprocessing.image import ImageDataGenerator
from keras.models import load_model
from keras.callbacks import ModelCheckpoint
from keras.utils import np_utils
import gc
args = parse_input()
data_dir = args.data_dir
validation_dir = args.validation_dir
model_file = args.model
save_dir = args.save_dir
tf_ordering = True
if (keras.backend.image_dim_ordering() == "th"):
tf_ordering = False
print("Image ordering:", keras.backend.image_dim_ordering(), "tf_ordering",
tf_ordering)
use_N4Correction = False
print("Using N4 correction", use_N4Correction)
training_samples = 1000
from inputParser import parse_input
import math
points = parse_input()
# Sort points by x
points.sort(key=lambda point: point[0])
def euclidean_distance(p1, p2):
return math.sqrt((p1[0] - p2[0])**2 + (p1[1] - p2[1])**2)
def x_belt_predicate(p1, p2, dist):
return p1[0] - p2[0] < dist
def y_belt_predicate(p1, p2, dist):
return p1[1] - p2[1] < dist
def get_min_dist_in_points(points):
if (len(points) <= 1):
return float('inf')
minimum = float('inf')
for i in range(0, len(points)):
for i2 in range(i + 1, len(points)):
distance = euclidean_distance(points[i], points[i2])
if distance < minimum:
minimum = distance