diff --git a/weclone/core/PII/pii_detector.py b/weclone/core/PII/pii_detector.py
index a2c376d..e210783 100644
--- a/weclone/core/PII/pii_detector.py
+++ b/weclone/core/PII/pii_detector.py
@@ -1,7 +1,7 @@
 from dataclasses import dataclass
 from typing import List, Optional, cast
 
-from presidio_analyzer import AnalyzerEngine, Pattern, PatternRecognizer
+from presidio_analyzer import AnalyzerEngine, Pattern, PatternRecognizer, RecognizerRegistry
 from presidio_analyzer.nlp_engine import NlpEngineProvider
 from presidio_anonymizer import AnonymizerEngine
 from presidio_anonymizer.entities.engine.recognizer_result import (
@@ -179,7 +179,9 @@ class ChinesePIIDetector(PIIDetector):
     """Chinese PII detector, extended to recognize Chinese-specific PII"""
 
     def __init__(self, threshold: float = 0.5):
-        super().__init__(language="zh", threshold=threshold)
+        self.language = "zh"
+        self.threshold = threshold
+        self._init_engines()
 
         # Filter out country-specific entities that are not relevant for Chinese context
         country_prefixes = ["US_", "UK_", "SG_", "AU_", "IN_"]
@@ -196,6 +198,12 @@ class ChinesePIIDetector(PIIDetector):
         ]
         logger.info(f"Chinese PII filtered entity types: {self.filtered_entities}")
 
+    def _init_engines(self):
+        registry = RecognizerRegistry()
+        self.analyzer = AnalyzerEngine(registry=registry)
+
+        self._add_custom_recognizers(language="zh")
+
     def _add_custom_recognizers(self, language: str):
         # Add parent class recognizers first
         super()._add_custom_recognizers(language="zh")