Merge pull request #195 from LDOUBLEV/fixocr

discard export PYTHONPATH manually and fix east label map dynamic shape

README.md

@@ -69,11 +69,6 @@ cd ..
 
 ```bash
 
-# 设置PYTHONPATH环境变量
-export PYTHONPATH=.
-# windows下设置环境变量
-SET PYTHONPATH=.
-
 # 预测image_dir指定的单张图像
 python3 tools/infer/predict_system.py --image_dir="./doc/imgs/11.jpg" --det_model_dir="./inference/ch_det_mv3_db/"  --rec_model_dir="./inference/ch_rec_mv3_crnn/"
 

README_en.md

@@ -69,11 +69,6 @@ The following code implements text detection and recognition inference tandemly.
 
 ```bash
 
-# Set PYTHONPATH environment variable
-export PYTHONPATH=.
-# Setting environment variable in Windows
-SET PYTHONPATH=.
-
 # Prediction on a single image by specifying image path to image_dir
 python3 tools/infer/predict_system.py --image_dir="./doc/imgs/11.jpg" --det_model_dir="./inference/ch_det_mv3_db/"  --rec_model_dir="./inference/ch_rec_mv3_crnn/"
 

ppocr/modeling/architectures/det_model.py

@@ -59,16 +59,23 @@ class DetModel(object):
         return: (image, corresponding label, dataloader)
         """
         image_shape = deepcopy(self.image_shape)
+        if image_shape[1] % 4 != 0 or image_shape[2] % 4 != 0:
+            raise Exception("The size of the image must be divisible by 4, "
+                            "received image shape is {}, please reset the "
+                            "Global.image_shape in the yml file".format(
+                                image_shape))
+
         image = fluid.layers.data(
             name='image', shape=image_shape, dtype='float32')
         if mode == "train":
             if self.algorithm == "EAST":
+                h, w = int(image_shape[1] // 4), int(image_shape[2] // 4)
                 score = fluid.layers.data(
-                    name='score', shape=[1, 128, 128], dtype='float32')
+                    name='score', shape=[1, h, w], dtype='float32')
                 geo = fluid.layers.data(
-                    name='geo', shape=[9, 128, 128], dtype='float32')
+                    name='geo', shape=[9, h, w], dtype='float32')
                 mask = fluid.layers.data(
-                    name='mask', shape=[1, 128, 128], dtype='float32')
+                    name='mask', shape=[1, h, w], dtype='float32')
                 feed_list = [image, score, geo, mask]
                 labels = {'score': score, 'geo': geo, 'mask': mask}
             elif self.algorithm == "DB":

tools/eval.py

@@ -17,6 +17,10 @@ from __future__ import division
 from __future__ import print_function
 
 import os
+import sys
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..'))
 
 
 def set_paddle_flags(**kwargs):

tools/export_model.py

@@ -18,9 +18,9 @@ from __future__ import print_function
 
 import os
 import sys
-import time
-import multiprocessing
-import numpy as np
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..'))
 
 
 def set_paddle_flags(**kwargs):

tools/infer/predict_det.py

@@ -11,8 +11,13 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+import os
+import sys
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '../..'))
 
-import utility
+import tools.infer.utility as utility
 from ppocr.utils.utility import initial_logger
 logger = initial_logger()
 from ppocr.utils.utility import get_image_file_list

tools/infer/predict_rec.py

@@ -11,8 +11,13 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
+import os
+import sys
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '../..'))
 
-import utility
+import tools.infer.utility as utility
 from ppocr.utils.utility import initial_logger
 logger = initial_logger()
 from ppocr.utils.utility import get_image_file_list

tools/infer/predict_system.py

@@ -16,12 +16,13 @@ import sys
 __dir__ = os.path.dirname(__file__)
 sys.path.append(__dir__)
 sys.path.append(os.path.join(__dir__, '../..'))
-import utility
+
+import tools.infer.utility as utility
 from ppocr.utils.utility import initial_logger
 logger = initial_logger()
 import cv2
-import predict_det
-import predict_rec
+import tools.infer.predict_det as predict_det
+import tools.infer.predict_rec as predict_rec
 import copy
 import numpy as np
 import math

tools/infer_det.py

@@ -16,14 +16,15 @@ from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
 
-import os
-import sys
-import time
 import numpy as np
 from copy import deepcopy
 import json
 
-# from paddle.fluid.contrib.model_stat import summary
+import os
+import sys
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..'))
 
 
 def set_paddle_flags(**kwargs):

tools/infer_rec.py

@@ -16,10 +16,12 @@ from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
 
-import os
-import time
-import multiprocessing
 import numpy as np
+import os
+import sys
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..'))
 
 
 def set_paddle_flags(**kwargs):
@@ -35,10 +37,7 @@ set_paddle_flags(
     FLAGS_eager_delete_tensor_gb=0,  # enable GC to save memory
 )
 
-from paddle import fluid
-
-# from ppocr.utils.utility import load_config, merge_config
-import program
+import tools.program as program
 from paddle import fluid
 from ppocr.utils.utility import initial_logger
 logger = initial_logger()
@@ -47,7 +46,6 @@ from ppocr.utils.save_load import init_model
 from ppocr.utils.character import CharacterOps
 from ppocr.utils.utility import create_module
 from ppocr.utils.utility import get_image_file_list
-logger = initial_logger()
 
 
 def main():

tools/train.py

@@ -18,9 +18,9 @@ from __future__ import print_function
 
 import os
 import sys
-import time
-import multiprocessing
-import numpy as np
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..'))
 
 
 def set_paddle_flags(**kwargs):