feat: enhance heading level feature with conditional imports and error handling

mineru/backend/vlm/token_to_middle_json.py

 import time
-import cv2
-import numpy as np
 from loguru import logger
-
-from mineru.backend.pipeline.model_init import AtomModelSingleton
+import numpy as np
+import cv2
 from mineru.utils.config_reader import get_llm_aided_config
 from mineru.utils.cut_image import cut_image_and_table
 from mineru.utils.enum_class import ContentType
 from mineru.utils.hash_utils import str_md5
 from mineru.backend.vlm.vlm_magic_model import MagicModel
-from mineru.utils.llm_aided import llm_aided_title
 from mineru.utils.pdf_image_tools import get_crop_img
 from mineru.version import __version__
 
+heading_level_import_success = False
+try:
+    from mineru.utils.llm_aided import llm_aided_title
+    from mineru.backend.pipeline.model_init import AtomModelSingleton
+    heading_level_import_success = True
+except Exception as e:
+    logger.warning("The heading level feature cannot be used. If you need to use the heading level feature, "
+                   "please execute `pip install mineru[pipeline]` to install the required packages.")
+
 
 def token_to_page_info(token, image_dict, page, image_writer, page_index) -> dict:
     """将token转换为页面信息"""
@@ -37,6 +43,7 @@ def token_to_page_info(token, image_dict, page, image_writer, page_index) -> dic
         title_aided_config = llm_aided_config.get('title_aided', None)
         if title_aided_config is not None:
             if title_aided_config.get('enable', False):
+                if heading_level_import_success:
                     atom_model_manager = AtomModelSingleton()
                     ocr_model = atom_model_manager.get_atom_model(
                         atom_model_name='ocr',
@@ -86,12 +93,12 @@ def result_to_middle_json(token_list, images_list, pdf_doc, image_writer):
 
     """llm优化"""
     llm_aided_config = get_llm_aided_config()
-
     if llm_aided_config is not None:
         """标题优化"""
         title_aided_config = llm_aided_config.get('title_aided', None)
         if title_aided_config is not None:
             if title_aided_config.get('enable', False):
+                if heading_level_import_success:
                     llm_aided_title_start_time = time.time()
                     llm_aided_title(middle_json["pdf_info"], title_aided_config)
                     logger.info(f'llm aided title time: {round(time.time() - llm_aided_title_start_time, 2)}')

pyproject.toml

@@ -33,6 +33,8 @@ dependencies = [
     "modelscope>=1.26.0",
     "huggingface-hub>=0.32.4",
     "json-repair>=0.46.2",
+    "opencv-python>=4.11.0.86",
+    "fast-langdetect>=0.2.3,<0.3.0",
 ]
 
 [project.optional-dependencies]
@@ -60,7 +62,6 @@ pipeline = [
     "torch>=2.2.2,!=2.5.0,!=2.5.1,<3",
     "torchvision",
     "transformers>=4.49.0,!=4.51.0,<5.0.0",
-    "fast-langdetect>=0.2.3,<0.3.0",
 ]
 api = [
     "fastapi",
@@ -97,7 +98,6 @@ pipeline_old_linux = [
     "torch>=2.2.2,!=2.5.0,!=2.5.1,<3",
     "torchvision",
     "transformers>=4.49.0,!=4.51.0,<5.0.0",
-    "fast-langdetect>=0.2.3,<0.3.0",
 ]
 
 [project.urls]