def get_saved_classifier():
'''Getting classifier from saved data if possible or training/generating new'''
data_dir = util.get_data_dir()
if not os.path.exists(data_dir):
os.mkdir(data_dir)
#Trying to load saved neural network coefficients
if os.path.exists(data_dir + "/" + config.network_name + ".npy"):
input_layer = config.sample_h * config.sample_w
outer_layer = config.character_number
neural_classifier = neural.NeuralClassfier(input_layer,
config.hidden_layer,
outer_layer,
config.reg,
random_seed=config.seed)
neural_classifier.weights = numpy.load(data_dir + "/" +
config.network_name + ".npy")
return neural_classifier
#Trying to load saved training data
if (os.path.exists(data_dir + "/X.npy")
and os.path.exists(data_dir + "/y.npy")):
X = numpy.load(data_dir + "/X.npy")
y = numpy.load(data_dir + "/y.npy")
else:
image_dir = util.get_image_dir()
if not os.path.exists(image_dir):
os.mkdir(image_dir)
if len(glob.glob(image_dir + "/*.gif")) < config.gen_train_size / 2:
debug("Generating recognized captcha images using perl script.")
generate_captcha(config.gen_train_size)
debug(
"Saved training data is not found. Generating new by segmentating images."
)
X, y = segment.extract_features()
numpy.save(data_dir + "/X.npy", X)
numpy.save(data_dir + "/y.npy", y)
debug("Network coefficients are not found. Training new neural network.")
neural_classifier = neural.train_network(X, y)
numpy.save(data_dir + "/" + config.network_name + ".npy",
neural_classifier.weights)
debug("Selfchecking full captcha files.")
accuracy = test.check_labeled_dir(neural_classifier,
util.get_image_dir(),
limit=100)
debug("Accuracy on generated set: {}".format(accuracy))
return neural_classifier
def get_buildfile_path(path: str, image_name: str) -> str:
'''
Returns the path of the buildfile.
:param path: The path of the image directory.
'''
image_dir = util.get_image_dir(path, image_name)
buildfile = os.path.join(image_dir, default.Config.BUILDFILE_NAME.value)
if not os.path.isfile(buildfile):
raise ValueError(f'buildfile does not exist: {buildfile}')
return buildfile
def build(
self,
namespace: argparse.Namespace,
image: str,
build_config: config.ImageBuildConfig):
'''
:param namespace: Namespace passed in via CLI.
:param image: The image to build.
:param build_config: the image build configuration.
:raises: subprocess.CalledProcessError
'''
LOGGER.info('Build image')
LOGGER.info('Dry Run: %s', namespace.dry_run)
image_dir = util.get_image_dir(namespace.path, image)
image = build_config.image.full_name
build_args = builder.get_build_args(build_config)
command = util.Command(['buildah', 'bud', '-t'])
command.add_arg(image)
command.add_args_list('--build-arg', build_args)
version = build_config.image.tag_build.version
if version:
command.add_args('--build-arg', f'VERSION={version}')
# add build context
command.add_arg('.')
LOGGER.info('Image name: %s', image)
LOGGER.info('Command: %s', ' '.join(command))
if namespace.dry_run:
return
# build
with util.pushd(image_dir):
subprocess.check_call(command)
image_tag = build_config.image.tag
if namespace.tag_latest and image_tag != 'latest':
self.tag_latest(image)
if namespace.push:
self.push(
namespace=namespace,
image=image)
def extract_features():
'''Extract features from all labeled images.'''
image_dir = util.get_image_dir()
images = glob.glob(image_dir+"/*.gif")
characters = []
var_segments = []
def extract_single(image_file):
captcha = re.match("(.*)\.gif",os.path.basename(image_file)).group(1)
image = util.read_grey_image(image_file)
return image_to_features(image,captcha)
X,y = list(zip(*list(map(extract_single,images))))
#return X,y
X = numpy.concatenate(X,axis=0)
y = numpy.concatenate(y,axis=0)
return (X,y)
def build(self, namespace: argparse.Namespace, image: str,
build_config: config.ImageBuildConfig):
'''
:param namespace: Namespace passed in via CLI.
:param image: The image to build.
:param build_config: the image build configuration.
:raises: subprocess.CalledProcessError
'''
LOGGER.info('Build image')
image_dir = util.get_image_dir(namespace.path, image)
image = build_config.image.full_name
build_args = builder.get_build_args(build_config)
version = build_config.image.tag_build.version
if version:
build_args.append(f'VERSION={version}')
command = self._create_command(namespace=namespace,
action='build',
build_args=build_args)
image_names_output = [image]
image_tag = build_config.image.tag
if namespace.tag_latest and image_tag != 'latest':
image_latest = util.set_image_tag_latest(image=image)
image_names_output.append(image_latest)
names_output = ','.join([f'name={i}' for i in image_names_output])
command.add_args(
name='--output',
value=f'type=image,{names_output},push={namespace.push}')
LOGGER.info('Image name: %s', image)
LOGGER.info('Command: %s', ' '.join(command))
if namespace.dry_run:
return
with util.pushd(image_dir):
subprocess.check_call(command)
def check_labeled_dir(NN, dir, limit=None, shift=0):
'''Checking accuracy on <captcha>.gif files.'''
total = 0
recognized = 0
if not os.path.exists(dir):
os.mkdir(dir)
images = glob.glob(dir + "/*.gif")
for image_file in images[shift:]:
total += 1
captcha_p = wkcaptcha.predict_file(NN, image_file)
captcha = re.match("(.*)\.gif", os.path.basename(image_file)).group(1)
if (captcha == captcha_p):
recognized += 1
if (limit and total >= limit):
break
return recognized / total
if __name__ == '__main__':
if (len(sys.argv) > 1):
debug("Accuracy on captcha in {} directory: {}".format(
sys.argv[1],
check_labeled_dir(wkcaptcha.get_saved_classifier(), sys.argv[1])))
else:
debug("Accuracy on generated set: {}".format(
check_labeled_dir(wkcaptcha.get_saved_classifier(),
util.get_image_dir(),
limit=100)))