def find_repeat_img(data_pool, data_dir, repeat_dir):
"""
查询重复的图片
:param data_pool:图片池
:param data_dir: 待处理图片
:param repeat_dir: 重复图片保存目录
:return:
"""
pool = []
for root, dirs, files in os.walk(data_pool):
for file in files:
name, _ = file.split(".")
if name not in pool:
pool.append(name)
if not os.path.exists(repeat_dir):
os.makedirs(repeat_dir)
total = 0
repeat = 0
for root, dirs, files in os.walk(data_dir):
for file in files:
name, _ = file.split(".")
total += 1
if name in pool:
repeat += 1
file_util.move_file(os.path.join(root, file), repeat_dir)
print("total={0}, repeat={1}".format(total, repeat))
pass
def verify_img(img_dir, error_img_save_dir):
"""
图片校验，删除有问题的图片
:param img_dir:
:param error_img_save_dir:
:return:
"""
total = 0
remove = 0
if not os.path.exists(error_img_save_dir):
os.makedirs(error_img_save_dir)
for root, dirs, files in os.walk(img_dir):
for img in files:
if not img_util.is_img(img):
continue
total += 1
img_file = os.path.join(root, img)
image = cv2.imread(img_file, cv2.IMREAD_GRAYSCALE)
if image is None:
print(img_file)
file_util.move_file(img_file, error_img_save_dir)
remove += 1
print("共{0}张图片，有问题的图片{1}张".format(total, remove))
def move_limit_width_height(img_dir,
out_limit_dir,
limit_width,
limit_height,
ext_list={'.JPG', '.jpg', '.png'}):
"""
将img_dir中所有宽或高大于限制宽或高的图片移动到out_limit_dir目录中
:param img_dir:
:param out_limit_dir:
:param limit_width:
:param limit_height:
:param ext_list:
:return:
"""
if not os.path.exists(out_limit_dir):
os.makedirs(out_limit_dir)
total = 0
limit = 0
for root, dirs, files in os.walk(img_dir):
for f in files:
ext = file_util.file_extension(os.path.join(root, f))
if ext in ext_list:
total += 1
img = cv2.imread(os.path.join(root, f))
if img is None:
file_util.move_file(os.path.join(root, f), out_limit_dir)
json_f = file_util.file_basename(f)
json_f += '.json'
file_util.move_file(os.path.join(root, json_f),
out_limit_dir)
continue
sp = img.shape
height = sp[0]
width = sp[1]
channel = sp[2]
if width > limit_width:
file_util.move_file(os.path.join(root, f), out_limit_dir)
json_f = file_util.file_basename(f)
json_f += '.json'
file_util.move_file(os.path.join(root, json_f),
out_limit_dir)
print('width: %d height: %d channel: %d' %
(width, height, channel))
limit += 1
continue
if height > limit_height:
file_util.move_file(os.path.join(root, f), out_limit_dir)
json_f = file_util.file_basename(f)
json_f += '.json'
file_util.move_file(os.path.join(root, json_f),
out_limit_dir)
print('width: %d height: %d channel: %d' %
(width, height, channel))
limit += 1
continue
print('total=' + str(total) + ',limit=' + str(limit))