Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into dygraph_rc

ppocr/postprocess/rec_postprocess.py

@@ -60,7 +60,7 @@ class BaseRecLabelDecode(object):
     def add_special_char(self, dict_character):
         return dict_character
 
-    def decode(self, text_index, text_prob=None, is_remove_duplicate=True):
+    def decode(self, text_index, text_prob=None, is_remove_duplicate=False):
         """ convert text-index into text-label. """
         result_list = []
         ignored_tokens = self.get_ignored_tokens()
@@ -110,7 +110,7 @@ class CTCLabelDecode(BaseRecLabelDecode):
         text = self.decode(preds_idx, preds_prob)
         if label is None:
             return text
-        label = self.decode(label, is_remove_duplicate=False)
+        label = self.decode(label)
         return text, label
 
     def add_special_char(self, dict_character):

ppocr/utils/utility.py

@@ -57,7 +57,7 @@ def get_image_file_list(img_file):
     elif os.path.isdir(img_file):
         for single_file in os.listdir(img_file):
             file_path = os.path.join(img_file, single_file)
-            if imghdr.what(file_path) in img_end:
+            if os.path.isfile(file_path) and imghdr.what(file_path) in img_end:
                 imgs_lists.append(file_path)
     if len(imgs_lists) == 0:
         raise Exception("not found any img file in {}".format(img_file))