update model

app.py

@@ -1,13 +1,10 @@
 import os, io
-from paddleocr import PaddleOCR, draw_ocr
+from paddleocr import PaddleOCR, draw_ocr,PPStructure
 from PIL import Image, ImageDraw
 import gradio as gr
 
 
-# 設定 Hugging Face Hub 的 Access Token
-os.environ["HF_TOKEN"] = "TWOCR"
- 
-def inference(img_path):
+def inference__ppocr(img_path):
 
     ocr = PaddleOCR(
         rec_char_dict_path='zhtw_common_dict.txt',
@@ -30,14 +27,36 @@ def inference(img_path):
     im_show_pil = draw_ocr(image, boxes, txts, scores, font_path="./simfang.ttf")
 
     return im_show_pil, "\n".join(txts)
-    
 
 
+def inference__ppstructure(img_path):
+
+    ppsutructure = PPStructure(
+        rec_char_dict_path='zhtw_common_dict.txt',
+        use_gpu=False,
+        rec_image_shape="3, 48, 320",
+        ser_dict_path='ppocr/utils/dict/kie/clinical_class_list.txt'
+    )
+    
+    result = ppsutructure.__call__(img_path)
+
+    for idx in range(len(result)):
+        res = result[idx]
+        for line in res:
+            print(line)
+    
+    result = result[0]
+    image = Image.open(img_path).convert('RGB')
+    boxes = [line[0] for line in result]
+    txts = [line[1][0] if line[1] else '' for line in result]  # 確保在無文字時 txts 還是個空字串
+    scores = [line[1][1] for line in result]
+    im_show_pil = draw_ocr(image, boxes, txts, scores, font_path="./simfang.ttf")
 
-css = ".output_image, .input_image {height: 40rem !important; width: 100% !important;}"
+    return im_show_pil, "\n".join(txts)
 
+    
 gr.Interface(
-    inference,
+    inference__ppstructure,
     [gr.Image(type='filepath', label='圖片上傳')],
     outputs=[
         gr.Image(type="pil", label="識別結果"),

paddleocr.py

@@ -286,30 +286,18 @@ MODEL_URLS = {
             }
         },
         'PP-StructureV2': {
-            'table': {
+            'kie': {
                 'en': {
                     'url':
-                    'https://paddleocr.bj.bcebos.com/ppstructure/models/slanet/en_ppstructure_mobile_v2.0_SLANet_infer.tar',
-                    'dict_path': 'ppocr/utils/dict/table_structure_dict.txt'
-                },
-                'ch': {
-                    'url':
-                    'https://paddleocr.bj.bcebos.com/ppstructure/models/slanet/ch_ppstructure_mobile_v2.0_SLANet_infer.tar',
-                    'dict_path': 'ppocr/utils/dict/table_structure_dict_ch.txt'
-                }
-            },
-            'layout': {
-                'en': {
-                    'url':
-                    'https://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_infer.tar',
+                    'https://huggingface.co/spaces/CallMeMrFern/ocr/ppstructure/models/kie/ser_clinical.tar',
                     'dict_path':
-                    'ppocr/utils/dict/layout_dict/layout_publaynet_dict.txt'
+                    'ppocr/utils/dict/kie/clinical_class_list.txt'
                 },
-                'ch': {
+                'tw': {
                     'url':
-                    'https://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_cdla_infer.tar',
+                    'https://huggingface.co/spaces/CallMeMrFern/ocr/ppstructure/models/kie/ser_clinical.tar',
                     'dict_path':
-                    'ppocr/utils/dict/layout_dict/layout_cdla_dict.txt'
+                    'ppocr/utils/dict/kie/clinical_class_list.txt'
                 }
             }
         }
@@ -346,7 +334,7 @@ def parse_args(mMain=True):
 
     for action in parser._actions:
         if action.dest in [
-                'rec_char_dict_path', 'table_char_dict_path', 'layout_dict_path'
+                'rec_char_dict_path', 'table_char_dict_path', 'layout_dict_path','kie_dict_path'
         ]:
             action.default = None
     if mMain:
@@ -409,6 +397,7 @@ def get_model_config(type, version, model_type, lang):
     model_urls = MODEL_URLS[type]
     if version not in model_urls:
         version = DEFAULT_MODEL_VERSION
+
     if model_type not in model_urls[version]:
         if model_type in model_urls[DEFAULT_MODEL_VERSION]:
             version = DEFAULT_MODEL_VERSION
@@ -416,7 +405,6 @@ def get_model_config(type, version, model_type, lang):
             logger.error('{} models is not support, we only support {}'.format(
                 model_type, model_urls[DEFAULT_MODEL_VERSION].keys()))
             sys.exit(-1)
-
     if lang not in model_urls[version][model_type]:
         if lang in model_urls[DEFAULT_MODEL_VERSION][model_type]:
             version = DEFAULT_MODEL_VERSION
@@ -607,31 +595,44 @@ class PPStructure(StructureSystem):
         params.rec_model_dir, rec_url = confirm_model_dir_url(
             params.rec_model_dir,
             os.path.join(BASE_DIR, 'whl', 'rec', lang), rec_model_config['url'])
-        table_model_config = get_model_config(
-            'STRUCTURE', params.structure_version, 'table', table_lang)
-        params.table_model_dir, table_url = confirm_model_dir_url(
-            params.table_model_dir,
-            os.path.join(BASE_DIR, 'whl', 'table'), table_model_config['url'])
-        layout_model_config = get_model_config(
-            'STRUCTURE', params.structure_version, 'layout', lang)
-        params.layout_model_dir, layout_url = confirm_model_dir_url(
-            params.layout_model_dir,
-            os.path.join(BASE_DIR, 'whl', 'layout'), layout_model_config['url'])
+        
+        # table_model_config = get_model_config(
+        #     'STRUCTURE', params.structure_version, 'table', table_lang)
+        # params.table_model_dir, table_url = confirm_model_dir_url(
+        #     params.table_model_dir,
+        #     os.path.join(BASE_DIR, 'whl', 'table'), table_model_config['url'])
+        # print(params.structure_version)
+        # layout_model_config = get_model_config(
+        #     'STRUCTURE', params.structure_version, 'layout', lang)
+        # params.layout_model_dir, layout_url = confirm_model_dir_url(
+        #     params.layout_model_dir,
+        #     os.path.join(BASE_DIR, 'whl', 'layout'), layout_model_config['url'])
+        
+        ser_model_config = get_model_config(
+            'STRUCTURE', params.structure_version, 'kie', table_lang)
+        params.ser_model_dir, ser_url = confirm_model_dir_url(
+            params.ser_model_dir,
+            os.path.join(BASE_DIR, 'whl', 'kie'), ser_model_config['url'])
+        print(params.ser_model_dir)
         # download model
         maybe_download(params.det_model_dir, det_url)
         maybe_download(params.rec_model_dir, rec_url)
-        maybe_download(params.table_model_dir, table_url)
-        maybe_download(params.layout_model_dir, layout_url)
+        # maybe_download(params.table_model_dir, table_url)
+        # maybe_download(params.layout_model_dir, layout_url)
+        maybe_download(params.ser_model_dir, ser_url)
 
         if params.rec_char_dict_path is None:
             params.rec_char_dict_path = str(
                 Path(__file__).parent / rec_model_config['dict_path'])
-        if params.table_char_dict_path is None:
-            params.table_char_dict_path = str(
-                Path(__file__).parent / table_model_config['dict_path'])
-        if params.layout_dict_path is None:
-            params.layout_dict_path = str(
-                Path(__file__).parent / layout_model_config['dict_path'])
+        # if params.table_char_dict_path is None:
+        #     params.table_char_dict_path = str(
+        #         Path(__file__).parent / table_model_config['dict_path'])
+        # if params.layout_dict_path is None:
+        #     params.layout_dict_path = str(
+        #         Path(__file__).parent / layout_model_config['dict_path'])
+        if params.ser_dict_path is None:
+            params.ser_dict_path = str(
+                Path(__file__).parent / ser_model_config['dict_path'])
         logger.debug(params)
         super().__init__(params)
 

ppocr/utils/dict/kie/clinical_class_list.txt

@@ -0,0 +1,13 @@
+hospitalname_key
+hospitalname_value
+hospitaladdress_key
+hospitaladdress_value
+department_key
+department_value
+diagnose_key
+diagnose_value
+doctorcomment_key
+doctorcomment_value
+doctorname_key
+doctorname_value
+other

ppstructure/models/kie/ser_clinical.tar

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f2f54efb1b23b0e6ac00238c7d5060f6a285ff06d0e2e5d0c30ce24e7fd0fff
+size 1113108480

ppstructure/utility.py

@@ -65,7 +65,7 @@ def init_args():
         "--mode",
         type=str,
         choices=['structure', 'kie'],
-        default='structure',
+        default='kie',
         help='structure and kie is supported')
     parser.add_argument(
         "--image_orientation",

test.json

@@ -0,0 +1 @@
+{"table": {"en": {"url": "https://paddleocr.bj.bcebos.com/ppstructure/models/slanet/en_ppstructure_mobile_v2.0_SLANet_infer.tar", "dict_path": "ppocr/utils/dict/table_structure_dict.txt"}, "ch": {"url": "https://paddleocr.bj.bcebos.com/ppstructure/models/slanet/ch_ppstructure_mobile_v2.0_SLANet_infer.tar", "dict_path": "ppocr/utils/dict/table_structure_dict_ch.txt"}}, "layout": {"en": {"url": "https://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_infer.tar", "dict_path": "ppocr/utils/dict/layout_dict/layout_publaynet_dict.txt"}, "ch": {"url": "https://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_cdla_infer.tar", "dict_path": "ppocr/utils/dict/layout_dict/layout_cdla_dict.txt"}}, "kie": {"en": {"url": "https://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_infer.tar", "dict_path": "ppocr/utils/dict/kie/clinical_class_list.txt"}, "ch": {"url": "https://huggingface.co/spaces/CallMeMrFern/ocr/ppstructure/models/kie/ser_clinical.tar", "dict_path": "ppocr/utils/dict/kie/clinical_class_list.txt"}}}