format_into_folder.py

#Run only once to sort out data
#Glauc Test Data images = ['G-39-L.jpg', 'S-29-R.jpg', 'G-33-R.jpg', 'S-16-R.jpg', 'G-24-L.jpg', 'S-25-L.jpg', 'S-31-L.jpg', 'S-5-L.jpg', 'S-18-L.jpg', 'G-13-R.jpg', 'G-18-R.jpg', 'G-35-R.jpg', 'G-7-L.jpg', 'S-6-R.jpg']
#Health Test Data images = ['N-47-L.jpg', 'N-70-R.jpg', 'N-58-R.jpg', 'N-11-L.jpg', 'N-22-R.jpg', 'N-69-L.jpg', 'N-13-L.jpg', 'N-81-L.jpg', 'N-6-R.jpg', 'N-21-L.jpg', 'N-18-R.jpg', 'N-44-R.jpg', 'N-3-L.jpg', 'N-90-R.jpg', 'N-23-L.jpg', 'N-55-L.jpg', 'N-32-R.jpg']
import os, random
from shutil import copy
from dotenv import load_dotenv
from PIL import Image
load_dotenv(".config.env")
glauc_path = os.getenv("GLAUC_PATH")
health_path = os.getenv("HEALTH_PATH")

dst = os.getenv("FORMATTED_DATA_PATH")

glauc_dir = os.listdir(glauc_path)
health_dir = os.listdir(health_path)


glauc_train = random.sample(glauc_dir, len(glauc_dir))
health_train = random.sample(health_dir, len(health_dir))

glauc_test = random.sample(glauc_dir, int(len(glauc_dir)/5))
health_test = random.sample(health_dir, int(len(health_dir)/5))

glauc_train = list(set(glauc_train).difference(set(glauc_test)))
health_train = list(set(health_train).difference(set(health_test)))

def crop(image_path, loc):
    image_obj = Image.open(image_path)
    cropped_image = image_obj.crop((0,0,image_obj.size[0]/2,image_obj.size[1]))
    cropped_image.save(loc)

for x in glauc_train:
    crop(glauc_path + '/' + x, dst + '/train/glaucoma/' + x)
    #copy(glauc_path + '/' + x, dst + '/train/glaucoma/')

for x in glauc_test:
    crop(glauc_path + '/' + x, dst + '/validation/glaucoma/' + x)
    #copy(glauc_path + '/' + x, dst + '/validation/glaucoma')

for x in health_train:
    crop(health_path + '/' + x, dst + '/train/healthy/' + x)
    #copy(health_path + '/' + x, dst + '/train/healthy')

for x in health_test:
    crop(health_path + '/' + x, dst + '/validation/healthy/' + x)
    #copy(health_path + '/' + x, dst + '/validation/healthy')