"""A multi-thread tool to crop large images to sub-images for faster IO."""
import os
import os.path as osp
import numpy as np
import cv2
from PIL import Image
import data.util as data_util # noqa: E402
import lmdb
import pyarrow
import torch.utils.data as data
from tqdm import tqdm
def main():
mode = 'single' # single (one input folder) | pair (extract corresponding GT and LR pairs)
split_img = False
opt = {}
opt['n_thread'] = 12
opt['compression_level'] = 90 # JPEG compression quality rating.
# CV_IMWRITE_PNG_COMPRESSION from 0 to 9. A higher value means a smaller size and longer
# compression time. If read raw images during training, use 0 for faster IO speed.
if mode == 'single':
opt['input_folder'] = 'F:\\4k6k\\datasets\\ns_images\\imagesets\\images'
opt['save_folder'] = 'F:\\4k6k\\datasets\\ns_images\\imagesets\\lmdb_with_ref'
opt['crop_sz'] = 512 # the size of each sub-image
opt['step'] = 128 # step of the sliding crop window
opt['thres_sz'] = 128 # size threshold
opt['resize_final_img'] = .5
opt['only_resize'] = False
extract_single(opt, split_img)
elif mode == 'pair':
GT_folder = '../../datasets/div2k/DIV2K_train_HR'
LR_folder = '../../datasets/div2k/DIV2K_train_LR_bicubic/X4'
save_GT_folder = '../../datasets/div2k/DIV2K800_sub'
save_LR_folder = '../../datasets/div2k/DIV2K800_sub_bicLRx4'
scale_ratio = 4
crop_sz = 480 # the size of each sub-image (GT)
step = 240 # step of the sliding crop window (GT)
thres_sz = 48 # size threshold
########################################################################
# check that all the GT and LR images have correct scale ratio
img_GT_list = data_util._get_paths_from_images(GT_folder)
img_LR_list = data_util._get_paths_from_images(LR_folder)
assert len(img_GT_list) == len(img_LR_list), 'different length of GT_folder and LR_folder.'
for path_GT, path_LR in zip(img_GT_list, img_LR_list):
img_GT = Image.open(path_GT)
img_LR = Image.open(path_LR)
w_GT, h_GT = img_GT.size
w_LR, h_LR = img_LR.size
assert w_GT / w_LR == scale_ratio, 'GT width [{:d}] is not {:d}X as LR weight [{:d}] for {:s}.'.format( # noqa: E501
w_GT, scale_ratio, w_LR, path_GT)
assert w_GT / w_LR == scale_ratio, 'GT width [{:d}] is not {:d}X as LR weight [{:d}] for {:s}.'.format( # noqa: E501
w_GT, scale_ratio, w_LR, path_GT)
# check crop size, step and threshold size
assert crop_sz % scale_ratio == 0, 'crop size is not {:d}X multiplication.'.format(
scale_ratio)
assert step % scale_ratio == 0, 'step is not {:d}X multiplication.'.format(scale_ratio)
assert thres_sz % scale_ratio == 0, 'thres_sz is not {:d}X multiplication.'.format(
scale_ratio)
print('process GT...')
opt['input_folder'] = GT_folder
opt['save_folder'] = save_GT_folder
opt['crop_sz'] = crop_sz
opt['step'] = step
opt['thres_sz'] = thres_sz
extract_single(opt)
print('process LR...')
opt['input_folder'] = LR_folder
opt['save_folder'] = save_LR_folder
opt['crop_sz'] = crop_sz // scale_ratio
opt['step'] = step // scale_ratio
opt['thres_sz'] = thres_sz // scale_ratio
extract_single(opt)
assert len(data_util._get_paths_from_images(save_GT_folder)) == len(
data_util._get_paths_from_images(
save_LR_folder)), 'different length of save_GT_folder and save_LR_folder.'
else:
raise ValueError('Wrong mode.')
class LmdbWriter:
def __init__(self, lmdb_path, max_mem_size=30*1024*1024*1024, write_freq=5000):
self.db = lmdb.open(lmdb_path, subdir=True,
map_size=max_mem_size, readonly=False,
meminit=False, map_async=True)
self.txn = self.db.begin(write=True)
self.ref_id = 0
self.tile_ids = {}
self.writes = 0
self.write_freq = write_freq
self.keys = []
# Writes the given reference image to the db and returns its ID.
def write_reference_image(self, ref_img):
id = self.ref_id
self.ref_id += 1
self.write_image(id, ref_img[0], ref_img[1])
return id
# Writes a tile image to the db given a reference image and returns its ID.
def write_tile_image(self, ref_id, tile_image):
next_tile_id = 0 if ref_id not in self.tile_ids.keys() else self.tile_ids[ref_id]
self.tile_ids[ref_id] = next_tile_id+1
full_id = "%i_%i" % (ref_id, next_tile_id)
self.write_image(full_id, tile_image[0], tile_image[1])
self.keys.append(full_id)
return full_id
# Writes an image directly to the db with the given reference image and center point.
def write_image(self, id, img, center_point):
self.txn.put(u'{}'.format(id).encode('ascii'), pyarrow.serialize(img).to_buffer(), pyarrow.serialize(center_point).to_buffer())
self.writes += 1
if self.writes % self.write_freq == 0:
self.txn.commit()
self.txn = self.db.begin(write=True)
def close(self):
self.txn.commit()
with self.db.begin(write=True) as txn:
txn.put(b'__keys__', pyarrow.serialize(self.keys).to_buffer())
txn.put(b'__len__', pyarrow.serialize(len(self.keys)).to_buffer())
self.db.sync()
self.db.close()
class TiledDataset(data.Dataset):
def __init__(self, opt, split_mode=False):
self.split_mode = split_mode
self.opt = opt
input_folder = opt['input_folder']
self.images = data_util._get_paths_from_images(input_folder)
def __getitem__(self, index):
if self.split_mode:
return self.get(index, True, True).extend(self.get(index, True, False))
else:
return self.get(index, False, False)
def get(self, index, split_mode, left_img):
path = self.images[index]
crop_sz = self.opt['crop_sz']
step = self.opt['step']
thres_sz = self.opt['thres_sz']
only_resize = self.opt['only_resize']
img = cv2.imread(path, cv2.IMREAD_UNCHANGED)
# We must convert the image into a square. Crop the image so that only the center is left, since this is often
# the most salient part of the image.
h, w, c = img.shape
dim = min(h, w)
img = img[(h - dim) // 2:dim + (h - dim) // 2, (w - dim) // 2:dim + (w - dim) // 2, :]
h, w, c = img.shape
# Uncomment to filter any image that doesnt meet a threshold size.
if min(h,w) < 1024:
return
left = 0
right = w
if split_mode:
if left_img:
left = 0
right = int(w/2)
else:
left = int(w/2)
right = w
w = int(w/2)
img = img[:, left:right]
h_space = np.arange(0, h - crop_sz + 1, step)
if h - (h_space[-1] + crop_sz) > thres_sz:
h_space = np.append(h_space, h - crop_sz)
w_space = np.arange(0, w - crop_sz + 1, step)
if w - (w_space[-1] + crop_sz) > thres_sz:
w_space = np.append(w_space, w - crop_sz)
dsize = None
if only_resize:
dsize = (crop_sz, crop_sz)
if h < w:
h_space = [0]
w_space = [(w - h) // 2]
crop_sz = h
else:
h_space = [(h - w) // 2]
w_space = [0]
crop_sz = w
index = 0
resize_factor = self.opt['resize_final_img'] if 'resize_final_img' in self.opt.keys() else 1
dsize = (int(crop_sz * resize_factor), int(crop_sz * resize_factor))
# Reference image should always be first.
results = [(cv2.resize(img, dsize, interpolation=cv2.INTER_AREA), (-1,-1))]
for x in h_space:
for y in w_space:
index += 1
crop_img = img[x:x + crop_sz, y:y + crop_sz, :]
center_point = (x + crop_sz // 2, y + crop_sz // 2)
crop_img = np.ascontiguousarray(crop_img)
if 'resize_final_img' in self.opt.keys():
# Resize too.
resize_factor = self.opt['resize_final_img']
center_point = (int(center_point[0] * resize_factor), int(center_point[1] * resize_factor))
crop_img = cv2.resize(crop_img, dsize, interpolation=cv2.INTER_AREA)
success, buffer = cv2.imencode(".jpg", crop_img, [cv2.IMWRITE_JPEG_QUALITY, self.opt['compression_level']])
assert success
results.append((buffer, center_point))
return results
def __len__(self):
return len(self.images)
def identity(x):
return x
def extract_single(opt, split_img=False):
save_folder = opt['save_folder']
if not osp.exists(save_folder):
os.makedirs(save_folder)
print('mkdir [{:s}] ...'.format(save_folder))
lmdb = LmdbWriter(save_folder)
dataset = TiledDataset(opt, split_img)
dataloader = data.DataLoader(dataset, num_workers=opt['n_thread'], collate_fn=identity)
tq = tqdm(dataloader)
for imgs in tq:
if imgs is None or len(imgs) <= 1:
continue
ref_id = lmdb.write_reference_image(imgs[0])
for tile in imgs[1:]:
lmdb.write_tile_image(ref_id, tile)
lmdb.close()
if __name__ == '__main__':
main()