def calculate_the_best_weight(self):
if os.listdir(checkpoint_path):
value = Image_Processing.extraction_image(checkpoint_path)
extract_num = [
os.path.splitext(os.path.split(i)[-1])[0] for i in value
]
num = [re.split('-', i) for i in extract_num]
accs = [float(i[-1]) for i in num]
losses = [float('-' + str(abs(float(i[-2])))) for i in num]
index = [acc + loss for acc, loss in zip(accs, losses)]
model_dict = dict((ind, val) for ind, val in zip(index, value))
return model_dict.get(max(index))
else:
logger.debug('没有可用的检查点')
from works.project_12306_ocr.settings import train_path from works.project_12306_ocr.Function_API import Image_Processing Image_Processing.rename_suffix(Image_Processing.extraction_image(train_path))
from works.project_12306_ocr.Callback import CallBack
from works.project_12306_ocr.settings import MODEL
from works.project_12306_ocr.settings import MODEL_NAME
from works.project_12306_ocr.settings import checkpoint_path
from works.project_12306_ocr.settings import App_model_path
from works.project_12306_ocr.settings import MULITI_MODEL_PREDICTION
from works.project_12306_ocr.Function_API import Distinguish_image
from works.project_12306_ocr.Function_API import Image_Processing
gpus = tf.config.experimental.list_physical_devices(device_type='GPU')
for gpu in gpus:
tf.config.experimental.set_memory_growth(device=gpu, enable=True)
app = Flask(__name__)
if os.listdir(App_model_path):
model_path = Image_Processing.extraction_image(App_model_path)
logger.debug(f'{model_path}模型加载成功')
else:
model = operator.methodcaller(MODEL)(Models)
try:
model.load_weights(os.path.join(checkpoint_path, CallBack.calculate_the_best_weight()))
except:
raise OSError(f'没有任何的权重和模型在{App_model_path}')
model_save = os.path.join(App_model_path, MODEL_NAME)
model.save(model_save)
model_path = [model_save]
logger.debug(f'{model_path}模型加载成功')
# logger.debug(model_path)
import random from works.project_12306_ocr.settings import train_path from works.project_12306_ocr.Function_API import Image_Processing train_image = Image_Processing.extraction_image(train_path) random.shuffle(train_image) Image_Processing.move_path(train_image)
from works.project_12306_ocr.settings import DATA_ENHANCEMENT
from works.project_12306_ocr.settings import train_path
from works.project_12306_ocr.settings import train_pack_path
from works.project_12306_ocr.settings import validation_pack_path
from works.project_12306_ocr.settings import test_pack_path
from works.project_12306_ocr.settings import train_enhance_path
from works.project_12306_ocr.Function_API import cheak_path
from works.project_12306_ocr.Function_API import Image_Processing
from works.project_12306_ocr.Function_API import parse_function_ocr
gpus = tf.config.experimental.list_physical_devices(device_type='GPU')
for gpu in gpus:
tf.config.experimental.set_memory_growth(device=gpu, enable=True)
train_dataset = tf.data.TFRecordDataset(
Image_Processing.extraction_image(train_pack_path)).map(
parse_function_ocr).batch(BATCH_SIZE)
validation_dataset = tf.data.TFRecordDataset(
Image_Processing.extraction_image(validation_pack_path)).map(
parse_function_ocr).batch(BATCH_SIZE)
test_dataset = tf.data.TFRecordDataset(
Image_Processing.extraction_image(test_pack_path)).map(
parse_function_ocr).batch(BATCH_SIZE)
logger.debug(train_dataset)
model, c_callback = CallBack.callback(operator.methodcaller(MODEL)(Models))
model.summary()
from works.project_12306_ocr.settings import train_path from works.project_12306_ocr.Function_API import Image_Processing Image_Processing.rename_path(train_path)
from works.project_12306_ocr.settings import validation_path
from works.project_12306_ocr.settings import test_path
from works.project_12306_ocr.settings import IMAGE_HEIGHT
from works.project_12306_ocr.settings import IMAGE_WIDTH
from works.project_12306_ocr.settings import train_enhance_path
from works.project_12306_ocr.settings import DATA_ENHANCEMENT
from works.project_12306_ocr.settings import TFRecord_train_path
from works.project_12306_ocr.settings import TFRecord_validation_path
from works.project_12306_ocr.settings import TFRecord_test_path
from works.project_12306_ocr.Function_API import Image_Processing
from works.project_12306_ocr.Function_API import WriteTFRecord
from concurrent.futures import ThreadPoolExecutor
if DATA_ENHANCEMENT:
with ThreadPoolExecutor(max_workers=100) as t:
for i in Image_Processing.extraction_image(train_path):
task = t.submit(Image_Processing.preprosess_save_images, i,
[IMAGE_HEIGHT, IMAGE_WIDTH])
train_image = Image_Processing.extraction_image(train_enhance_path)
Image_Processing.save_ocr_class(train_image)
random.shuffle(train_image)
train_lable = Image_Processing.extraction_ocr_lable(train_image)
else:
train_image = Image_Processing.extraction_image(train_path)
Image_Processing.save_ocr_class(train_image)
random.shuffle(train_image)
train_lable = Image_Processing.extraction_ocr_lable(train_image)
validation_image = Image_Processing.extraction_image(validation_path)
validation_lable = Image_Processing.extraction_ocr_lable(validation_image)
from works.project_12306_ocr.settings import test_path
from works.project_12306_ocr.settings import model_path
from works.project_12306_ocr.settings import BATCH_SIZE
from works.project_12306_ocr.settings import test_pack_path
from works.project_12306_ocr.settings import MODEL_LEAD_NAME
from works.project_12306_ocr.settings import MULITI_MODEL_PREDICTION
from works.project_12306_ocr.Function_API import Image_Processing
from works.project_12306_ocr.Function_API import Distinguish_image
from works.project_12306_ocr.Function_API import parse_function_verification
gpus = tf.config.experimental.list_physical_devices(device_type='GPU')
for gpu in gpus:
tf.config.experimental.set_memory_growth(device=gpu, enable=True)
test_dataset = tf.data.TFRecordDataset(
Image_Processing.extraction_image(test_pack_path)).map(
parse_function_verification).batch(BATCH_SIZE)
test_image_list = Image_Processing.extraction_image(test_path)
random.shuffle(test_image_list)
if MULITI_MODEL_PREDICTION:
model_path = Image_Processing.extraction_image(model_path)
if not model_path:
raise OSError(f'{model_path}没有模型')
for i in test_image_list[:10]:
Distinguish_image.distinguish_images(model_path, i)
else:
model_path = os.path.join(model_path, MODEL_LEAD_NAME)
logger.debug(f'加载模型{model_path}')
if not os.path.exists(model_path):