Feat: add support for the Ascend layout recognizer (#10105)

### What problem does this PR solve?

Supports Ascend layout recognizer.

Use the environment variable `LAYOUT_RECOGNIZER_TYPE=ascend` to enable
the Ascend layout recognizer, and `ASCEND_LAYOUT_RECOGNIZER_DEVICE_ID=n`
(for example, n=0) to specify the Ascend device ID.

Ensure that you have installed the [ais
tools](https://gitee.com/ascend/tools/tree/master/ais-bench_workload/tool/ais_bench)
properly.

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

deepdoc/vision/__init__.py

@@ -16,24 +16,28 @@
 import io
 import sys
 import threading
+
 import pdfplumber
 
 from .ocr import OCR
 from .recognizer import Recognizer
+from .layout_recognizer import AscendLayoutRecognizer
 from .layout_recognizer import LayoutRecognizer4YOLOv10 as LayoutRecognizer
 from .table_structure_recognizer import TableStructureRecognizer
 
-
 LOCK_KEY_pdfplumber = "global_shared_lock_pdfplumber"
 if LOCK_KEY_pdfplumber not in sys.modules:
     sys.modules[LOCK_KEY_pdfplumber] = threading.Lock()
 
 
 def init_in_out(args):
-    from PIL import Image
     import os
     import traceback
+
+    from PIL import Image
+
     from api.utils.file_utils import traversal_files
+
     images = []
     outputs = []
 
@@ -44,8 +48,7 @@ def init_in_out(args):
         nonlocal outputs, images
         with sys.modules[LOCK_KEY_pdfplumber]:
             pdf = pdfplumber.open(fnm)
-            images = [p.to_image(resolution=72 * zoomin).annotated for i, p in
-                                enumerate(pdf.pages)]
+            images = [p.to_image(resolution=72 * zoomin).annotated for i, p in enumerate(pdf.pages)]
 
         for i, page in enumerate(images):
             outputs.append(os.path.split(fnm)[-1] + f"_{i}.jpg")
@@ -57,10 +60,10 @@ def init_in_out(args):
             pdf_pages(fnm)
             return
         try:
-            fp = open(fnm, 'rb')
+            fp = open(fnm, "rb")
             binary = fp.read()
             fp.close()
-            images.append(Image.open(io.BytesIO(binary)).convert('RGB'))
+            images.append(Image.open(io.BytesIO(binary)).convert("RGB"))
             outputs.append(os.path.split(fnm)[-1])
         except Exception:
             traceback.print_exc()
@@ -81,6 +84,7 @@ __all__ = [
     "OCR",
     "Recognizer",
     "LayoutRecognizer",
+    "AscendLayoutRecognizer",
     "TableStructureRecognizer",
     "init_in_out",
 ]