Sfoglia il codice sorgente

refactor(main_v2): update output directory paths and configuration settings

- Removed the 'coordinate_precision' setting from the default output configuration to streamline options.
- Updated various output directory paths for test inputs to reflect new structure and naming conventions.
- Adjusted the configuration settings for logging and page processing to enhance clarity and organization.
- Ensured consistency in the handling of input and output directories across different test cases.
zhch158_admin 2 giorni fa
parent
commit
da9d741745
1 ha cambiato i file con 17 aggiunte e 10 eliminazioni
  1. 17 10
      ocr_tools/universal_doc_parser/main_v2.py

+ 17 - 10
ocr_tools/universal_doc_parser/main_v2.py

@@ -117,7 +117,6 @@ def _get_default_output_config(debug: bool) -> dict:
         'draw_type_label': True,
         'draw_bbox_number': True,
         'save_enhanced_json': True,
-        'coordinate_precision': 2,
         'normalize_numbers': True,
         'merge_cross_page_tables': True,
     }
@@ -416,7 +415,7 @@ if __name__ == "__main__":
         default_config = {
             # 测试输入
             # "input": "/Users/zhch158/workspace/data/流水分析/康强_北京农村商业银行.pdf",
-            # "output_dir": "./output/康强_北京农村商业银行_bank_statement_v2",
+            # "output_dir": "./output/康强_北京农村商业银行/bank_statement_mineru_vl",
 
             # "input": "/Users/zhch158/workspace/repository.git/ocr_platform/ocr_tools/universal_doc_parser/tests/A用户_单元格扫描流水_page_002.png",
             # "output_dir": "./output/A用户_单元格扫描流水_bank_statement_wired_unet",
@@ -428,7 +427,7 @@ if __name__ == "__main__":
             # "input": "/Users/zhch158/workspace/repository.git/ocr_platform/ocr_tools/universal_doc_parser/tests/2023年度报告母公司_page_003.png",
             # "input": "/Users/zhch158/workspace/repository.git/ocr_platform/ocr_tools/universal_doc_parser/tests/2023年度报告母公司_page_003_270_skew(-0.4).png",
             # "input": "/Users/zhch158/workspace/data/流水分析/2023年度报告母公司.pdf",
-            # "output_dir": "./output/2023年度报告母公司/bank_statement_wired_unet",
+            # "output_dir": "/Users/zhch158/workspace/data/流水分析/2023年度报告母公司/bank_statement_wired_unet",
 
             # "input": "/Users/zhch158/workspace/data/流水分析/2023年度报告母公司.pdf",
             # "output_dir": "/Users/zhch158/workspace/data/流水分析/2023年度报告母公司/bank_statement_yusys_v2",
@@ -442,11 +441,17 @@ if __name__ == "__main__":
             # "output_dir": "./output/600916_中国黄金_2022年报/bank_statement_wired_unet",
 
 
-            # "input": "/Users/zhch158/workspace/repository.git/ocr_platform/ocr_tools/universal_doc_parser/tests/提取自赤峰黄金2023年报.pdf",
-            # "output_dir": "./output/提取自赤峰黄金2023年报/bank_statement_wired_unet",
+            # "input": "/Users/zhch158/workspace/data/流水分析/德_内蒙古银行照.pdf",
+            # "output_dir": "./output/德_内蒙古银行照/bank_statement_wired_unet",
+
+            "input": "/Users/zhch158/workspace/repository.git/ocr_platform/ocr_tools/universal_doc_parser/tests/提取自赤峰黄金2023年报.pdf",
+            "output_dir": "./output/提取自赤峰黄金2023年报/bank_statement_wired_unet",
+            # "input": "/Users/zhch158/workspace/data/流水分析/提取自赤峰黄金2023年报.pdf",
+            # "output_dir": "/Users/zhch158/workspace/data/流水分析/提取自赤峰黄金2023年报/bank_statement_wired_unet",
 
             # "input": "/Users/zhch158/workspace/data/流水分析/施博深.pdf",
-            # "output_dir": "/Users/zhch158/workspace/data/流水分析/施博深/bank_statement_yusys_v2",
+            # "output_dir": "/Users/zhch158/workspace/data/流水分析/施博深/bank_statement_wired_unet",
+            # "output_dir": "./output/施博深/bank_statement_smart_router",
 
             # "input": "/Users/zhch158/workspace/data/流水分析/施博深.wiredtable/施博深_page_020.png",
             # "output_dir": "./output/施博深/bank_statement_wired_unet",
@@ -454,11 +459,13 @@ if __name__ == "__main__":
             # "input": "/Users/zhch158/workspace/data/流水分析/施博深.wiredtable",
             # "output_dir": "/Users/zhch158/workspace/data/流水分析/施博深/bank_statement_wired_unet",
 
-            "input": "/Users/zhch158/workspace/data/流水分析/山西云集科技有限公司.pdf",
-            "output_dir": "/Users/zhch158/workspace/data/流水分析/山西云集科技有限公司/bank_statement_wired_unet",
+            # "input": "/Users/zhch158/workspace/data/流水分析/山西云集科技有限公司.pdf",
+            # "output_dir": "/Users/zhch158/workspace/data/流水分析/山西云集科技有限公司/bank_statement_wired_unet",
 
             # 配置文件
             "config": "./config/bank_statement_wired_unet.yaml",
+            # "config": "./config/bank_statement_smart_router.yaml",
+            # "config": "./config/bank_statement_mineru_vl.yaml",
             # "config": "./config/bank_statement_yusys_v2.yaml",
             # "config": "./config/bank_statement_paddle_vl.yaml",
             
@@ -466,7 +473,7 @@ if __name__ == "__main__":
             "scene": "bank_statement",
             
             # 页面范围(可选)
-            # "pages": "12,26,27",  # 只处理前1页
+            "pages": "7",  # 只处理前1页
             # "pages": "1-3,5,7-10",  # 处理指定页面
             # "pages": "83-109",  # 处理指定页面
 
@@ -479,7 +486,7 @@ if __name__ == "__main__":
             "log_level": "DEBUG",
 
             # 日志文件
-            "log_file": "./logs/bank_statement_wired_unet/process.log",
+            "log_file": "./output/logs/bank_statement_wired_unet/process.log",
         }
         
         # 构造参数