images_prepare.py

# -*- coding: utf-8 -*-
"""

@author: serdarhelli
"""

import os
import numpy as np
from PIL import Image
from zipfile import ZipFile
from natsort import natsorted


def convert_one_channel(img):
    #some images have 3 channels , although they are grayscale image
    if len(img.shape)>2:
        img=img[:,:,0]
        return img
    else:
        return img
    
def pre_images(resize_shape,path,include_zip):
    if include_zip==True:
        ZipFile(path+"/DentalPanoramicXrays.zip").extractall(path) 
        path=path+'/Images/'
    dirs=natsorted(os.listdir(path))
    sizes=np.zeros([len(dirs),2])
    images=img=Image.open(path+dirs[0])
    sizes[0,:]=images.size
    images=(images.resize((resize_shape),Image.ANTIALIAS))
    images=convert_one_channel(np.asarray(images))
    for i in range (1,len(dirs)):
        img=Image.open(path+dirs[i])
        sizes[i,:]=img.size
        img=img.resize((resize_shape),Image.ANTIALIAS)
        img=convert_one_channel(np.asarray(img))
        images=np.concatenate((images,img))
    images=np.reshape(images,(len(dirs),resize_shape[0],resize_shape[1],1))
    return images,sizes