Browse Source

feat: 调整相似度阈值至95,以提高OCR结果比较的准确性

zhch158_admin 1 month ago
parent
commit
dd6ccc3d06
1 changed files with 1 additions and 1 deletions
  1. 1 1
      compare_ocr_results.py

+ 1 - 1
compare_ocr_results.py

@@ -12,7 +12,7 @@ from fuzzywuzzy import fuzz
 class OCRResultComparator:
     def __init__(self):
         self.differences = []
-        self.similarity_threshold = 85
+        self.similarity_threshold = 95
         self.max_paragraph_window = 6
         self.table_comparison_mode = 'standard'  # 新增:表格比较模式
         self.header_similarity_threshold = 80  # 表头相似度阈值