Procházet zdrojové kódy

feat(更新处理器配置): 添加PaddleOCR处理器配置,更新日志目录,新增YUSYS本地处理器支持

zhch158_admin před 1 týdnem
rodič
revize
0d52ee2e18

+ 2 - 1
ocr_tools/ocr_batch/pdf_list.txt

@@ -15,5 +15,6 @@ B用户_扫描流水.pdf,bank_statement
 湛_平安银行图.pdf,bank_statement
 张_微信图.pdf,bank_statement
 朱_中信银行图.pdf,bank_statement
-韩_中国银行图.pdf,bank_statement
+韩_中国银行图.pdf,bank_statemen
+严_农业银行.pdf,bank_statement
 

+ 1 - 0
ocr_tools/ocr_batch/pdf_list_local.txt

@@ -5,3 +5,4 @@ B用户_扫描流水.pdf,bank_statement
 施博深.pdf,bank_statement
 山西云集科技有限公司.pdf,bank_statement
 许_民生银行图.pdf,bank_statement
+严_农业银行.pdf,bank_statement

+ 17 - 1
ocr_tools/ocr_batch/processor_configs.yaml

@@ -74,10 +74,26 @@ processors:
       - "--debug"
       - "--log_level=DEBUG"
     output_subdir: "bank_statement_yusys_local"
-    log_subdir: "logs/bank_statement_yusys_v4_local"
+    log_subdir: "logs/bank_statement_yusys_local"
     venv: "conda activate mineru2"
     description: "YUSYS(local) Wired UNET OCR GLM-OCR"
 
+  yusys_paddleocr_local:
+    script: "/Users/zhch158/workspace/repository.git/ocr_platform/ocr_tools/universal_doc_parser/main_v2.py"
+    input_arg: "--input"
+    output_arg: "--output_dir"
+    scene_arg: "--scene"
+    extra_args:
+      - "--config=/Users/zhch158/workspace/repository.git/ocr_platform/ocr_tools/universal_doc_parser/config/bank_statement_paddle_vl_local.yaml"
+      - "--pages=1-35"
+      - "--streaming"
+      - "--debug"
+      - "--log_level=DEBUG"
+    output_subdir: "bank_statement_yusys_paddleocr_local"
+    log_subdir: "logs/bank_statement_yusys_paddleocr_local"
+    venv: "conda activate mineru2"
+    description: "YUSYS(local) Wired UNET OCR PaddleOCR-VL"
+
   # -------------------------------------------------------------------------
   # PaddleOCR-VL 处理器
   # -------------------------------------------------------------------------