Browse Source

Merge pull request #769 from myhloli/add-doclayout-yolo

update:update config json
Xiaomeng Zhao 1 year ago
parent
commit
efb5851f4b
2 changed files with 13 additions and 4 deletions
  1. 12 3
      magic-pdf.template.json
  2. 1 1
      magic_pdf/dict2md/ocr_mkcontent.py

+ 12 - 3
magic-pdf.template.json

@@ -6,9 +6,18 @@
     "models-dir":"/tmp/models",
     "layoutreader-model-dir":"/tmp/layoutreader",
     "device-mode":"cpu",
+    "layout-config": {
+        "model": "doclayout_yolo"
+    },
+    "formula-config": {
+        "mfd_model": "yolo_v8_mfd",
+        "mfr_model": "unimernet_small",
+        "enable": true
+    },
     "table-config": {
-        "model": "TableMaster",
-        "is_table_recog_enable": false,
+        "model": "tablemaster",
+        "enable": false,
         "max_time": 400
-    }
+    },
+    "config_version": "1.0.0"
 }

+ 1 - 1
magic_pdf/dict2md/ocr_mkcontent.py

@@ -162,7 +162,7 @@ def merge_para_with_text(para_block):
 def para_to_standard_format_v2(para_block, img_buket_path, page_idx, drop_reason=None):
     para_type = para_block['type']
     para_content = {}
-    if para_type == BlockType.Text:
+    if para_type in [BlockType.Text, BlockType.List, BlockType.Index]:
         para_content = {
             'type': 'text',
             'text': merge_para_with_text(para_block),