refactor(vision): 优化姿态关键点检测器的初始化逻辑

- 移除未使用的 threading 和 time 模块导入
- 统一委托类型的使用，避免硬编码委托类型
- 简化 GPU 和 CPU 委托的创建流程
- 修复委托类型传递的一致性问题

app/vision/pose_landmarker.py

@@ -1,7 +1,5 @@
 from __future__ import annotations
 
-import threading
-import time
 from typing import Callable
 
 import mediapipe as mp
@@ -30,14 +28,14 @@ class PoseLandmarkerWrapper:
         if prefer_gpu:
             try:
                 self.delegate = BaseOptions.Delegate.GPU
-                self._landmarker = self._create(PoseLandmarker.Delegate.GPU)
+                self._landmarker = self._create(self.delegate, result_callback)
                 logger.info("MediaPipe PoseLandmarker initialized with GPU delegate")
                 return
             except Exception as exc:
                 logger.warning("MediaPipe GPU delegate unavailable, falling back to CPU: {}", exc)
 
         self.delegate = BaseOptions.Delegate.CPU
-        self._landmarker = self._create(PoseLandmarker.Delegate.CPU, result_callback)
+        self._landmarker = self._create(self.delegate, result_callback)
         logger.info("MediaPipe PoseLandmarker initialized with CPU delegate")
 
     def _create(self, delegate, result_callback=None):