.gitignore

@@ -13,3 +13,7 @@ no_frame_debug.log
 
 feeder/
 .venv/
+.vscode/
+dist/
+websocket_comm.log
+temp_debug/

Dockerfile.base

@@ -1,8 +1,17 @@
 # Base image with all ML dependencies
-FROM python:3.13-bookworm
+FROM pytorch/pytorch:2.8.0-cuda12.6-cudnn9-runtime
 
 # Install system dependencies
-RUN apt update && apt install -y libgl1 && rm -rf /var/lib/apt/lists/*
+RUN apt update && apt install -y \
+    libgl1 \
+    libglib2.0-0 \
+    libgstreamer1.0-0 \
+    libgtk-3-0 \
+    libavcodec58 \
+    libavformat58 \
+    libswscale5 \
+    libgomp1 \
+    && rm -rf /var/lib/apt/lists/*
 
 # Copy and install base requirements (ML dependencies that rarely change)
 COPY requirements.base.txt .

debug/test_camera_indices.py

@@ -0,0 +1,142 @@
+#!/usr/bin/env python3
+"""
+Test script to check available camera indices
+"""
+
+import cv2
+import logging
+import sys
+import subprocess
+
+# Configure logging
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s: %(message)s"
+)
+logger = logging.getLogger("camera_index_test")
+
+def test_camera_index(index):
+    """Test if a camera index is available"""
+    try:
+        cap = cv2.VideoCapture(index)
+        if cap.isOpened():
+            ret, frame = cap.read()
+            if ret and frame is not None:
+                width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+                height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+                fps = cap.get(cv2.CAP_PROP_FPS)
+                
+                cap.release()
+                return True, f"{width}x{height} @ {fps}fps"
+            else:
+                cap.release()
+                return False, "Can open but cannot read frames"
+        else:
+            cap.release()
+            return False, "Cannot open camera"
+    except Exception as e:
+        return False, f"Error: {str(e)}"
+
+def get_windows_cameras_ffmpeg():
+    """Get available cameras on Windows using FFmpeg"""
+    try:
+        result = subprocess.run(['ffmpeg', '-f', 'dshow', '-list_devices', 'true', '-i', 'dummy'], 
+                              capture_output=True, text=True, timeout=10, encoding='utf-8', errors='ignore')
+        output = result.stderr
+        
+        lines = output.split('\n')
+        video_devices = []
+        
+        # Parse the output - look for lines with (video) that contain device names in quotes
+        for line in lines:
+            if '[dshow @' in line and '(video)' in line and '"' in line:
+                # Extract device name between first pair of quotes
+                start = line.find('"') + 1
+                end = line.find('"', start)
+                if start > 0 and end > start:
+                    device_name = line[start:end]
+                    video_devices.append(device_name)
+        
+        logger.info(f"FFmpeg detected video devices: {video_devices}")
+        return video_devices
+    except Exception as e:
+        logger.error(f"Failed to get Windows camera names: {e}")
+        return []
+
+def main():
+    logger.info("=== Camera Index Test ===")
+    
+    # Check FFmpeg availability for Windows device detection
+    ffmpeg_available = False
+    try:
+        result = subprocess.run(['ffmpeg', '-version'], capture_output=True, text=True, timeout=5)
+        if result.returncode == 0:
+            ffmpeg_available = True
+            logger.info("FFmpeg is available")
+    except:
+        logger.info("FFmpeg not available")
+    
+    # Get Windows camera names if possible
+    if sys.platform.startswith('win') and ffmpeg_available:
+        logger.info("\n=== Windows Camera Devices (FFmpeg) ===")
+        cameras = get_windows_cameras_ffmpeg()
+        if cameras:
+            for i, camera in enumerate(cameras):
+                logger.info(f"Device {i}: {camera}")
+        else:
+            logger.info("No cameras detected via FFmpeg")
+    
+    # Test camera indices 0-9
+    logger.info("\n=== Testing Camera Indices ===")
+    available_cameras = []
+    
+    for index in range(10):
+        logger.info(f"Testing camera index {index}...")
+        is_available, info = test_camera_index(index)
+        
+        if is_available:
+            logger.info(f"✓ Camera {index}: AVAILABLE - {info}")
+            available_cameras.append(index)
+        else:
+            logger.info(f"✗ Camera {index}: NOT AVAILABLE - {info}")
+    
+    # Summary
+    logger.info("\n=== Summary ===")
+    if available_cameras:
+        logger.info(f"Available camera indices: {available_cameras}")
+        logger.info(f"Default camera index to use: {available_cameras[0]}")
+        
+        # Test the first available camera more thoroughly
+        logger.info(f"\n=== Detailed Test for Camera {available_cameras[0]} ===")
+        cap = cv2.VideoCapture(available_cameras[0])
+        if cap.isOpened():
+            # Get properties
+            width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+            height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+            fps = cap.get(cv2.CAP_PROP_FPS)
+            backend = cap.getBackendName()
+            
+            logger.info(f"Resolution: {width}x{height}")
+            logger.info(f"FPS: {fps}")
+            logger.info(f"Backend: {backend}")
+            
+            # Test frame capture
+            ret, frame = cap.read()
+            if ret and frame is not None:
+                logger.info(f"Frame capture: SUCCESS")
+                logger.info(f"Frame shape: {frame.shape}")
+                logger.info(f"Frame dtype: {frame.dtype}")
+            else:
+                logger.info(f"Frame capture: FAILED")
+            
+            cap.release()
+    else:
+        logger.error("No cameras available!")
+        logger.info("Possible solutions:")
+        logger.info("1. Check if camera is connected and not used by another application")
+        logger.info("2. Check camera permissions")
+        logger.info("3. Try different camera indices")
+        logger.info("4. Install camera drivers")
+
+if __name__ == "__main__":
+    main()

docs/WorkerConnection.md

@@ -2,7 +2,7 @@
 
 ## Overview
 
-The Camera Module implements a pure VMware DRS-like declarative architecture for managing connections to Python ML workers. This system uses the database as the single source of truth for desired subscription state, with automatic regeneration and reconciliation providing intelligent camera management, real-time object detection, and AI-powered content selection with automatic load balancing capabilities.
+The Camera Module implements a pure declarative architecture for managing connections to Python ML workers. This system uses the database as the single source of truth for desired subscription state, with automatic regeneration and reconciliation providing intelligent camera management, real-time object detection, and AI-powered content selection with automatic load balancing capabilities.
 
 **Key Architectural Principle**: Database mutations trigger complete state regeneration rather than incremental updates, ensuring consistency and eliminating complex state synchronization issues.
 
@@ -44,7 +44,7 @@ Core distributed cluster implementation that handles declarative state managemen
 **Master Mode Responsibilities:**
 - Maintains WebSocket connections to all Python workers
 - Manages desired vs actual subscription state separation
-- Implements VMware DRS-like global rebalancing algorithm
+- Implements intelligent global rebalancing algorithm
 - Processes automatic reconciliation every 30 seconds
 - Responds to slave join/leave events from MasterElection
 - Generates fresh pre-signed model URLs for worker assignments
@@ -201,10 +201,10 @@ All Redis data uses **manual cleanup only** (no TTL) to ensure:
 - Predictable cleanup during planned maintenance
 - Debug visibility into system state history
 
-## Pure Declarative Architecture (VMware DRS-like)
+## Pure Declarative Architecture
 
 ### Concept Overview
-The system implements a pure declarative approach similar to VMware Distributed Resource Scheduler (DRS), where:
+The system implements a pure declarative approach where:
 - **Database**: Single source of truth for desired state (Display+Camera+Playlist combinations)
 - **Actual State**: What subscriptions are currently running on workers (stored in `worker:actual_subscriptions`)
 - **Regeneration**: Master regenerates complete desired state from database on every change notification
@@ -261,8 +261,8 @@ async handleDatabaseChange(changeType: string, entityId: string) {
   }
 }
 
-// VMware DRS-like worker selection (unchanged)
-function findBestWorkerVMwareDRS(workers, currentLoads) {
+// Intelligent worker selection (unchanged)
+function findBestWorkerForLoad(workers, currentLoads) {
   return workers
     .map(worker => ({
       worker,
@@ -280,7 +280,7 @@ function findBestWorkerVMwareDRS(workers, currentLoads) {
 3. **Complete Regeneration**: Master queries database for all active Display+Camera+Playlist combinations
 4. **Desired State Creation**: Master generates fresh desired subscriptions from database query results
 5. **Diff Analysis**: Master compares fresh desired state vs current actual state on workers
-6. **Global Reconciliation**: Master applies VMware DRS algorithm to reconcile differences
+6. **Global Reconciliation**: Master applies intelligent load balancing algorithm to reconcile differences
 7. **Worker Commands**: Master sends subscription/unsubscription commands to workers
 8. **State Update**: Master updates actual subscription state in Redis
 
@@ -1483,7 +1483,7 @@ This interface specification provides external services with a clear understandi
 - **Service Layer Simplicity**: Services just update database + trigger regeneration - no subscription logic
 - **Operational Resilience**: System is self-healing and predictable - any database change triggers complete reconciliation
 
-### VMware DRS-like Benefits
+### Declarative Architecture Benefits
 - **Global Optimization**: Every regeneration considers all subscriptions globally for optimal load balancing
 - **Automatic Recovery**: System automatically heals from any inconsistent state by regenerating from database
 - **Resource Efficiency**: Workers assigned based on real-time CPU/memory metrics with load balancing
@@ -1495,4 +1495,4 @@ This interface specification provides external services with a clear understandi
 - **Memory Efficiency**: No persistent state storage outside of database and current worker assignments
 - **Network Efficiency**: Minimal command protocol reduces Redis pub/sub overhead
 
-This pure declarative architecture provides the reliability and simplicity of Kubernetes-style declarative resource management while maintaining the performance and scalability needed for real-time camera processing systems.
+This pure declarative architecture provides the reliability and simplicity of container orchestration-style declarative resource management while maintaining the performance and scalability needed for real-time camera processing systems.

requirements.base.txt

@@ -1,7 +1,12 @@
-torch
-torchvision
-ultralytics
-opencv-python
-scipy
-filterpy
-psycopg2-binary
+ultralytics>=8.3.0
+opencv-python>=4.6.0
+scipy>=1.9.0
+filterpy>=1.4.0
+psycopg2-binary>=2.9.0
+easydict
+loguru
+pyzmq
+gitpython
+gdown
+lap
+pynvml

requirements.txt

@@ -1,6 +1,5 @@
-fastapi
+fastapi[standard]
 uvicorn
 websockets
-fastapi[standard]
 redis
 urllib3<2.0.0

siwatsystem/database.py

@@ -80,37 +80,50 @@ class DatabaseManager:
         try:
             cur = self.connection.cursor()
             
-            # Build the UPDATE query dynamically
+            # Build the INSERT and UPDATE query dynamically
+            insert_placeholders = []
+            insert_values = [key_value]  # Start with key_value
+            
             set_clauses = []
-            values = []
+            update_values = []
             
             for field, value in fields.items():
                 if value == "NOW()":
+                    # Special handling for NOW()
+                    insert_placeholders.append("NOW()")
                     set_clauses.append(f"{field} = NOW()")
                 else:
+                    insert_placeholders.append("%s")
+                    insert_values.append(value)
                     set_clauses.append(f"{field} = %s")
-                    values.append(value)
+                    update_values.append(value)
             
             # Add schema prefix if table doesn't already have it
             full_table_name = table if '.' in table else f"gas_station_1.{table}"
             
+            # Build the complete query
             query = f"""
             INSERT INTO {full_table_name} ({key_field}, {', '.join(fields.keys())})
-            VALUES (%s, {', '.join(['%s'] * len(fields))})
+            VALUES (%s, {', '.join(insert_placeholders)})
             ON CONFLICT ({key_field})
             DO UPDATE SET {', '.join(set_clauses)}
             """
             
-            # Add key_value to the beginning of values list
-            all_values = [key_value] + list(fields.values()) + values
+            # Combine values for the query: insert_values + update_values
+            all_values = insert_values + update_values
+            
+            logger.debug(f"SQL Query: {query}")
+            logger.debug(f"Values: {all_values}")
             
             cur.execute(query, all_values)
             self.connection.commit()
             cur.close()
-            logger.info(f"Updated {table} for {key_field}={key_value}")
+            logger.info(f"✅ Updated {table} for {key_field}={key_value} with fields: {fields}")
             return True
         except Exception as e:
-            logger.error(f"Failed to execute update on {table}: {e}")
+            logger.error(f"❌ Failed to execute update on {table}: {e}")
+            logger.debug(f"Query: {query if 'query' in locals() else 'Query not built'}")
+            logger.debug(f"Values: {all_values if 'all_values' in locals() else 'Values not prepared'}")
             if self.connection:
                 self.connection.rollback()
             return False

siwatsystem/model_registry.py

@@ -0,0 +1,242 @@
+"""
+Shared Model Registry for Memory Optimization
+
+This module implements a global shared model registry to prevent duplicate model loading
+in memory when multiple cameras use the same model. This significantly reduces RAM and
+GPU VRAM usage by ensuring only one instance of each unique model is loaded.
+
+Key Features:
+- Thread-safe model loading and access
+- Reference counting for proper cleanup
+- Automatic model lifecycle management
+- Maintains compatibility with existing pipeline system
+"""
+
+import os
+import threading
+import logging
+from typing import Dict, Any, Optional, Set
+import torch
+from ultralytics import YOLO
+
+# Create a logger for this module
+logger = logging.getLogger("detector_worker.model_registry")
+
+class ModelRegistry:
+    """
+    Singleton class for managing shared YOLO models across multiple cameras.
+    
+    This registry ensures that each unique model is loaded only once in memory,
+    dramatically reducing RAM and GPU VRAM usage when multiple cameras use the
+    same model.
+    """
+    
+    _instance = None
+    _lock = threading.Lock()
+    
+    def __new__(cls):
+        if cls._instance is None:
+            with cls._lock:
+                if cls._instance is None:
+                    cls._instance = super(ModelRegistry, cls).__new__(cls)
+                    cls._instance._initialized = False
+        return cls._instance
+    
+    def __init__(self):
+        if self._initialized:
+            return
+            
+        self._initialized = True
+        
+        # Thread-safe storage for loaded models
+        self._models: Dict[str, YOLO] = {}  # modelId -> YOLO model instance
+        self._model_files: Dict[str, str] = {}  # modelId -> file path
+        self._reference_counts: Dict[str, int] = {}  # modelId -> reference count
+        self._model_lock = threading.RLock()  # Reentrant lock for nested calls
+        
+        logger.info("🏭 Shared Model Registry initialized - ready for memory-optimized model loading")
+    
+    def get_model(self, model_id: str, model_file_path: str) -> YOLO:
+        """
+        Get or load a YOLO model. Returns shared instance if already loaded.
+        
+        Args:
+            model_id: Unique identifier for the model
+            model_file_path: Path to the model file
+            
+        Returns:
+            YOLO model instance (shared across all callers)
+        """
+        with self._model_lock:
+            if model_id in self._models:
+                # Model already loaded - increment reference count and return
+                self._reference_counts[model_id] += 1
+                logger.info(f"📖 Model '{model_id}' reused (ref_count: {self._reference_counts[model_id]}) - SAVED MEMORY!")
+                return self._models[model_id]
+            
+            # Model not loaded yet - load it
+            logger.info(f"🔄 Loading NEW model '{model_id}' from {model_file_path}")
+            
+            if not os.path.exists(model_file_path):
+                raise FileNotFoundError(f"Model file {model_file_path} not found")
+            
+            try:
+                # Load the YOLO model
+                model = YOLO(model_file_path)
+                
+                # Move to GPU if available
+                if torch.cuda.is_available():
+                    logger.info(f"🚀 CUDA available. Moving model '{model_id}' to GPU VRAM")
+                    model.to("cuda")
+                else:
+                    logger.info(f"💻 CUDA not available. Using CPU for model '{model_id}'")
+                
+                # Store in registry
+                self._models[model_id] = model
+                self._model_files[model_id] = model_file_path
+                self._reference_counts[model_id] = 1
+                
+                logger.info(f"✅ Model '{model_id}' loaded and registered (ref_count: 1)")
+                self._log_registry_status()
+                
+                return model
+                
+            except Exception as e:
+                logger.error(f"❌ Failed to load model '{model_id}' from {model_file_path}: {e}")
+                raise
+    
+    def release_model(self, model_id: str) -> None:
+        """
+        Release a reference to a model. If reference count reaches zero,
+        the model may be unloaded to free memory.
+        
+        Args:
+            model_id: Unique identifier for the model to release
+        """
+        with self._model_lock:
+            if model_id not in self._reference_counts:
+                logger.warning(f"⚠️ Attempted to release unknown model '{model_id}'")
+                return
+            
+            self._reference_counts[model_id] -= 1
+            logger.info(f"📉 Model '{model_id}' reference count decreased to {self._reference_counts[model_id]}")
+            
+            # For now, keep models in memory even when ref count reaches 0
+            # This prevents reload overhead if the same model is needed again soon
+            # In the future, we could implement LRU eviction policy
+            # if self._reference_counts[model_id] <= 0:
+            #     logger.info(f"💤 Model '{model_id}' has 0 references but keeping in memory for reuse")
+                # Optionally: self._unload_model(model_id)
+    
+    def _unload_model(self, model_id: str) -> None:
+        """
+        Internal method to unload a model from memory.
+        Currently not used to prevent reload overhead.
+        """
+        with self._model_lock:
+            if model_id in self._models:
+                logger.info(f"🗑️ Unloading model '{model_id}' from memory")
+                
+                # Clear GPU memory if model was on GPU
+                model = self._models[model_id]
+                if hasattr(model, 'model') and hasattr(model.model, 'cuda'):
+                    try:
+                        # Move model to CPU before deletion to free GPU memory
+                        model.to('cpu')
+                    except Exception as e:
+                        logger.warning(f"⚠️ Failed to move model '{model_id}' to CPU: {e}")
+                
+                # Remove from registry
+                del self._models[model_id]
+                del self._model_files[model_id]
+                del self._reference_counts[model_id]
+                
+                # Force garbage collection
+                import gc
+                gc.collect()
+                if torch.cuda.is_available():
+                    torch.cuda.empty_cache()
+                
+                logger.info(f"✅ Model '{model_id}' unloaded and memory freed")
+                self._log_registry_status()
+    
+    def get_registry_status(self) -> Dict[str, Any]:
+        """
+        Get current status of the model registry.
+        
+        Returns:
+            Dictionary with registry statistics
+        """
+        with self._model_lock:
+            return {
+                "total_models": len(self._models),
+                "models": {
+                    model_id: {
+                        "file_path": self._model_files[model_id],
+                        "reference_count": self._reference_counts[model_id]
+                    }
+                    for model_id in self._models
+                },
+                "total_references": sum(self._reference_counts.values())
+            }
+    
+    def _log_registry_status(self) -> None:
+        """Log current registry status for debugging."""
+        status = self.get_registry_status()
+        logger.info(f"📊 Model Registry Status: {status['total_models']} unique models, {status['total_references']} total references")
+        for model_id, info in status['models'].items():
+            logger.debug(f"   📋 '{model_id}': refs={info['reference_count']}, file={os.path.basename(info['file_path'])}")
+    
+    def cleanup_all(self) -> None:
+        """
+        Clean up all models from the registry. Used during shutdown.
+        """
+        with self._model_lock:
+            model_ids = list(self._models.keys())
+            logger.info(f"🧹 Cleaning up {len(model_ids)} models from registry")
+            
+            for model_id in model_ids:
+                self._unload_model(model_id)
+            
+            logger.info("✅ Model registry cleanup complete")
+
+
+# Global singleton instance
+_registry = ModelRegistry()
+
+def get_shared_model(model_id: str, model_file_path: str) -> YOLO:
+    """
+    Convenience function to get a shared model instance.
+    
+    Args:
+        model_id: Unique identifier for the model
+        model_file_path: Path to the model file
+        
+    Returns:
+        YOLO model instance (shared across all callers)
+    """
+    return _registry.get_model(model_id, model_file_path)
+
+def release_shared_model(model_id: str) -> None:
+    """
+    Convenience function to release a shared model reference.
+    
+    Args:
+        model_id: Unique identifier for the model to release
+    """
+    _registry.release_model(model_id)
+
+def get_registry_status() -> Dict[str, Any]:
+    """
+    Convenience function to get registry status.
+    
+    Returns:
+        Dictionary with registry statistics
+    """
+    return _registry.get_registry_status()
+
+def cleanup_registry() -> None:
+    """
+    Convenience function to cleanup the entire registry.
+    """
+    _registry.cleanup_all()

siwatsystem/mpta_manager.py

@@ -0,0 +1,375 @@
+"""
+Shared MPTA Manager for Disk Space Optimization
+
+This module implements shared MPTA file management to prevent duplicate downloads
+and extractions when multiple cameras use the same model. MPTA files are stored
+in modelId-based directories and shared across all cameras using that model.
+
+Key Features:
+- Thread-safe MPTA downloading and extraction
+- ModelId-based directory structure: models/{modelId}/
+- Reference counting for proper cleanup
+- Eliminates duplicate MPTA downloads
+- Maintains compatibility with existing pipeline system
+"""
+
+import os
+import threading
+import logging
+import shutil
+import requests
+from typing import Dict, Set, Optional
+from urllib.parse import urlparse
+from .pympta import load_pipeline_from_zip
+
+# Create a logger for this module
+logger = logging.getLogger("detector_worker.mpta_manager")
+
+class MPTAManager:
+    """
+    Singleton class for managing shared MPTA files across multiple cameras.
+    
+    This manager ensures that each unique modelId is downloaded and extracted
+    only once, dramatically reducing disk usage and download time when multiple
+    cameras use the same model.
+    """
+    
+    _instance = None
+    _lock = threading.Lock()
+    
+    def __new__(cls):
+        if cls._instance is None:
+            with cls._lock:
+                if cls._instance is None:
+                    cls._instance = super(MPTAManager, cls).__new__(cls)
+                    cls._instance._initialized = False
+        return cls._instance
+    
+    def __init__(self):
+        if self._initialized:
+            return
+            
+        self._initialized = True
+        
+        # Thread-safe storage for MPTA management
+        self._model_paths: Dict[int, str] = {}  # modelId -> shared_extraction_path
+        self._mpta_file_paths: Dict[int, str] = {}  # modelId -> local_mpta_file_path
+        self._reference_counts: Dict[int, int] = {}  # modelId -> reference count  
+        self._download_locks: Dict[int, threading.Lock] = {}  # modelId -> download lock
+        self._cameras_using_model: Dict[int, Set[str]] = {}  # modelId -> set of camera_ids
+        self._manager_lock = threading.RLock()  # Reentrant lock for nested calls
+        
+        logger.info("🏭 Shared MPTA Manager initialized - ready for disk-optimized MPTA management")
+    
+    def get_or_download_mpta(self, model_id: int, model_url: str, camera_id: str) -> Optional[tuple[str, str]]:
+        """
+        Get or download an MPTA file. Returns (extraction_path, mpta_file_path) if successful.
+        
+        Args:
+            model_id: Unique identifier for the model
+            model_url: URL to download the MPTA file from
+            camera_id: Identifier for the requesting camera
+            
+        Returns:
+            Tuple of (extraction_path, mpta_file_path), or None if failed
+        """
+        with self._manager_lock:
+            # Track camera usage
+            if model_id not in self._cameras_using_model:
+                self._cameras_using_model[model_id] = set()
+            self._cameras_using_model[model_id].add(camera_id)
+            
+            # Check if model directory already exists on disk (from previous sessions)
+            if model_id not in self._model_paths:
+                potential_path = f"models/{model_id}"
+                if os.path.exists(potential_path) and os.path.isdir(potential_path):
+                    # Directory exists from previous session, find the MPTA file
+                    mpta_files = [f for f in os.listdir(potential_path) if f.endswith('.mpta')]
+                    if mpta_files:
+                        # Use the first .mpta file found
+                        mpta_file_path = os.path.join(potential_path, mpta_files[0])
+                        self._model_paths[model_id] = potential_path
+                        self._mpta_file_paths[model_id] = mpta_file_path
+                        self._reference_counts[model_id] = 0  # Will be incremented below
+                        logger.info(f"📂 Found existing MPTA modelId {model_id} from previous session")
+            
+            # Check if already available
+            if model_id in self._model_paths:
+                shared_path = self._model_paths[model_id]
+                mpta_file_path = self._mpta_file_paths.get(model_id)
+                if os.path.exists(shared_path) and mpta_file_path and os.path.exists(mpta_file_path):
+                    self._reference_counts[model_id] += 1
+                    logger.info(f"📂 MPTA modelId {model_id} reused for camera {camera_id} (ref_count: {self._reference_counts[model_id]}) - SAVED DOWNLOAD!")
+                    return (shared_path, mpta_file_path)
+                else:
+                    # Path was deleted externally, clean up our records
+                    logger.warning(f"⚠️ MPTA path for modelId {model_id} was deleted externally, will re-download")
+                    del self._model_paths[model_id]
+                    self._mpta_file_paths.pop(model_id, None)
+                    self._reference_counts.pop(model_id, 0)
+            
+            # Need to download - get or create download lock for this modelId
+            if model_id not in self._download_locks:
+                self._download_locks[model_id] = threading.Lock()
+            
+        # Download with model-specific lock (released _manager_lock to allow other models)
+        download_lock = self._download_locks[model_id]
+        with download_lock:
+            # Double-check after acquiring download lock
+            with self._manager_lock:
+                if model_id in self._model_paths and os.path.exists(self._model_paths[model_id]):
+                    mpta_file_path = self._mpta_file_paths.get(model_id)
+                    if mpta_file_path and os.path.exists(mpta_file_path):
+                        self._reference_counts[model_id] += 1
+                        logger.info(f"📂 MPTA modelId {model_id} became available during wait (ref_count: {self._reference_counts[model_id]})")
+                        return (self._model_paths[model_id], mpta_file_path)
+            
+            # Actually download and extract
+            shared_path = f"models/{model_id}"
+            logger.info(f"🔄 Downloading NEW MPTA for modelId {model_id} from {model_url}")
+            
+            try:
+                # Ensure directory exists
+                os.makedirs(shared_path, exist_ok=True)
+                
+                # Download MPTA file
+                mpta_filename = self._extract_filename_from_url(model_url) or f"model_{model_id}.mpta"
+                local_mpta_path = os.path.join(shared_path, mpta_filename)
+                
+                if not self._download_file(model_url, local_mpta_path):
+                    logger.error(f"❌ Failed to download MPTA for modelId {model_id}")
+                    return None
+                
+                # Extract MPTA
+                pipeline_tree = load_pipeline_from_zip(local_mpta_path, shared_path)
+                if pipeline_tree is None:
+                    logger.error(f"❌ Failed to extract MPTA for modelId {model_id}")
+                    return None
+                
+                # Success - register in manager
+                with self._manager_lock:
+                    self._model_paths[model_id] = shared_path
+                    self._mpta_file_paths[model_id] = local_mpta_path
+                    self._reference_counts[model_id] = 1
+                    
+                    logger.info(f"✅ MPTA modelId {model_id} downloaded and registered (ref_count: 1)")
+                    self._log_manager_status()
+                
+                return (shared_path, local_mpta_path)
+                
+            except Exception as e:
+                logger.error(f"❌ Error downloading/extracting MPTA for modelId {model_id}: {e}")
+                # Clean up partial download
+                if os.path.exists(shared_path):
+                    shutil.rmtree(shared_path, ignore_errors=True)
+                return None
+    
+    def release_mpta(self, model_id: int, camera_id: str) -> None:
+        """
+        Release a reference to an MPTA. If reference count reaches zero,
+        the MPTA directory may be cleaned up to free disk space.
+        
+        Args:
+            model_id: Unique identifier for the model to release
+            camera_id: Identifier for the camera releasing the reference
+        """
+        with self._manager_lock:
+            if model_id not in self._reference_counts:
+                logger.warning(f"⚠️ Attempted to release unknown MPTA modelId {model_id} for camera {camera_id}")
+                return
+            
+            # Remove camera from usage tracking
+            if model_id in self._cameras_using_model:
+                self._cameras_using_model[model_id].discard(camera_id)
+            
+            self._reference_counts[model_id] -= 1
+            logger.info(f"📉 MPTA modelId {model_id} reference count decreased to {self._reference_counts[model_id]} (released by {camera_id})")
+            
+            # Clean up if no more references
+            # if self._reference_counts[model_id] <= 0:
+            #     self._cleanup_mpta(model_id)
+    
+    def _cleanup_mpta(self, model_id: int) -> None:
+        """
+        Internal method to clean up an MPTA directory and free disk space.
+        """
+        if model_id in self._model_paths:
+            shared_path = self._model_paths[model_id]
+            
+            try:
+                if os.path.exists(shared_path):
+                    shutil.rmtree(shared_path)
+                    logger.info(f"🗑️ Cleaned up MPTA directory: {shared_path}")
+                
+                # Remove from tracking
+                del self._model_paths[model_id]
+                self._mpta_file_paths.pop(model_id, None)
+                del self._reference_counts[model_id]
+                self._cameras_using_model.pop(model_id, None)
+                
+                # Clean up download lock (optional, could keep for future use)
+                self._download_locks.pop(model_id, None)
+                
+                logger.info(f"✅ MPTA modelId {model_id} fully cleaned up and disk space freed")
+                self._log_manager_status()
+                
+            except Exception as e:
+                logger.error(f"❌ Error cleaning up MPTA modelId {model_id}: {e}")
+    
+    def get_shared_path(self, model_id: int) -> Optional[str]:
+        """
+        Get the shared extraction path for a modelId without downloading.
+        
+        Args:
+            model_id: Model identifier to look up
+            
+        Returns:
+            Shared path if available, None otherwise
+        """
+        with self._manager_lock:
+            return self._model_paths.get(model_id)
+    
+    def get_manager_status(self) -> Dict:
+        """
+        Get current status of the MPTA manager.
+        
+        Returns:
+            Dictionary with manager statistics
+        """
+        with self._manager_lock:
+            return {
+                "total_mpta_models": len(self._model_paths),
+                "models": {
+                    str(model_id): {
+                        "shared_path": path,
+                        "reference_count": self._reference_counts.get(model_id, 0),
+                        "cameras_using": list(self._cameras_using_model.get(model_id, set()))
+                    }
+                    for model_id, path in self._model_paths.items()
+                },
+                "total_references": sum(self._reference_counts.values()),
+                "active_downloads": len(self._download_locks)
+            }
+    
+    def _log_manager_status(self) -> None:
+        """Log current manager status for debugging."""
+        status = self.get_manager_status()
+        logger.info(f"📊 MPTA Manager Status: {status['total_mpta_models']} unique models, {status['total_references']} total references")
+        for model_id, info in status['models'].items():
+            cameras_str = ','.join(info['cameras_using'][:3])  # Show first 3 cameras
+            if len(info['cameras_using']) > 3:
+                cameras_str += f"+{len(info['cameras_using'])-3} more"
+            logger.debug(f"   📋 ModelId {model_id}: refs={info['reference_count']}, cameras=[{cameras_str}]")
+    
+    def cleanup_all(self) -> None:
+        """
+        Clean up all MPTA directories. Used during shutdown.
+        """
+        with self._manager_lock:
+            model_ids = list(self._model_paths.keys())
+            logger.info(f"🧹 Cleaning up {len(model_ids)} MPTA directories")
+            
+            for model_id in model_ids:
+                self._cleanup_mpta(model_id)
+            
+            # Clear all tracking data
+            self._download_locks.clear()
+            logger.info("✅ MPTA manager cleanup complete")
+    
+    def _download_file(self, url: str, local_path: str) -> bool:
+        """
+        Download a file from URL to local path with progress logging.
+        
+        Args:
+            url: URL to download from
+            local_path: Local path to save to
+            
+        Returns:
+            True if successful, False otherwise
+        """
+        try:
+            logger.info(f"⬇️ Starting download from {url}")
+            
+            response = requests.get(url, stream=True)
+            response.raise_for_status()
+            
+            total_size = int(response.headers.get('content-length', 0))
+            if total_size > 0:
+                logger.info(f"📦 File size: {total_size / 1024 / 1024:.2f} MB")
+            
+            downloaded = 0
+            last_logged_progress = 0
+            with open(local_path, 'wb') as f:
+                for chunk in response.iter_content(chunk_size=8192):
+                    if chunk:
+                        f.write(chunk)
+                        downloaded += len(chunk)
+                        
+                        if total_size > 0:
+                            progress = int((downloaded / total_size) * 100)
+                            # Log at 10% intervals (10%, 20%, 30%, etc.)
+                            if progress >= last_logged_progress + 10 and progress <= 100:
+                                logger.debug(f"Download progress: {progress}%")
+                                last_logged_progress = progress
+            
+            logger.info(f"✅ Successfully downloaded to {local_path}")
+            return True
+            
+        except Exception as e:
+            logger.error(f"❌ Download failed: {e}")
+            # Clean up partial file
+            if os.path.exists(local_path):
+                os.remove(local_path)
+            return False
+    
+    def _extract_filename_from_url(self, url: str) -> Optional[str]:
+        """Extract filename from URL."""
+        try:
+            parsed = urlparse(url)
+            filename = os.path.basename(parsed.path)
+            return filename if filename else None
+        except Exception:
+            return None
+
+
+# Global singleton instance
+_mpta_manager = MPTAManager()
+
+def get_or_download_mpta(model_id: int, model_url: str, camera_id: str) -> Optional[tuple[str, str]]:
+    """
+    Convenience function to get or download a shared MPTA.
+    
+    Args:
+        model_id: Unique identifier for the model
+        model_url: URL to download the MPTA file from
+        camera_id: Identifier for the requesting camera
+        
+    Returns:
+        Tuple of (extraction_path, mpta_file_path), or None if failed
+    """
+    return _mpta_manager.get_or_download_mpta(model_id, model_url, camera_id)
+
+def release_mpta(model_id: int, camera_id: str) -> None:
+    """
+    Convenience function to release a shared MPTA reference.
+    
+    Args:
+        model_id: Unique identifier for the model to release
+        camera_id: Identifier for the camera releasing the reference
+    """
+    _mpta_manager.release_mpta(model_id, camera_id)
+
+def get_mpta_manager_status() -> Dict:
+    """
+    Convenience function to get MPTA manager status.
+    
+    Returns:
+        Dictionary with manager statistics
+    """
+    return _mpta_manager.get_manager_status()
+
+def cleanup_mpta_manager() -> None:
+    """
+    Convenience function to cleanup the entire MPTA manager.
+    """
+    _mpta_manager.cleanup_all()

test/test.py

@@ -0,0 +1,60 @@
+from ultralytics import YOLO
+import cv2
+import os
+
+# Load the model
+# model = YOLO('../models/webcam-local-01/4/bangchak_poc/yolo11n.pt')
+model = YOLO('yolo11m.pt')
+
+def test_image(image_path):
+    """Test a single image with YOLO model"""
+    if not os.path.exists(image_path):
+        print(f"Image not found: {image_path}")
+        return
+    
+    # Run inference - filter for car class only (class 2 in COCO)
+    results = model(image_path, classes=[2, 5, 7])  # 2, 5, 7 = car, bus, truck in COCO dataset
+    
+    # Display results
+    for r in results:
+        im_array = r.plot()  # plot a BGR numpy array of predictions
+        
+        # Resize image for display (max width/height 800px)
+        height, width = im_array.shape[:2]
+        max_dimension = 800
+        if width > max_dimension or height > max_dimension:
+            if width > height:
+                new_width = max_dimension
+                new_height = int(height * (max_dimension / width))
+            else:
+                new_height = max_dimension
+                new_width = int(width * (max_dimension / height))
+            im_array = cv2.resize(im_array, (new_width, new_height))
+        
+        # Show image with predictions
+        cv2.imshow('YOLO Test - Car Detection Only', im_array)
+        cv2.waitKey(0)
+        cv2.destroyAllWindows()
+        
+        # Print detection info
+        print(f"\nDetections for {image_path}:")
+        if r.boxes is not None and len(r.boxes) > 0:
+            for i, box in enumerate(r.boxes):
+                cls = int(box.cls[0])
+                conf = float(box.conf[0])
+                original_class = model.names[cls]  # Original class name (car/bus/truck)
+                # Get bounding box coordinates
+                x1, y1, x2, y2 = box.xyxy[0].tolist()
+                # Rename all vehicle types to "car"
+                print(f"Detection {i+1}: car (was: {original_class}) - Confidence: {conf:.3f} - BBox: ({x1:.0f}, {y1:.0f}, {x2:.0f}, {y2:.0f})")
+            print(f"Total cars detected: {len(r.boxes)}")
+        else:
+            print("No cars detected in the image")
+
+if __name__ == "__main__":
+    # Test with an image file
+    image_path = input("Enter image path (or press Enter for default test): ")
+    if not image_path:
+        image_path = "sample.png"  # Default test image
+    
+    test_image(image_path)

test/test_botsort_zone_track.py

@@ -0,0 +1,352 @@
+import cv2
+import torch
+import numpy as np
+import time
+from collections import defaultdict
+from ultralytics import YOLO
+
+def point_in_polygon(point, polygon):
+    """Check if a point is inside a polygon using ray casting algorithm"""
+    x, y = point
+    n = len(polygon)
+    inside = False
+    
+    p1x, p1y = polygon[0]
+    for i in range(1, n + 1):
+        p2x, p2y = polygon[i % n]
+        if y > min(p1y, p2y):
+            if y <= max(p1y, p2y):
+                if x <= max(p1x, p2x):
+                    if p1y != p2y:
+                        xinters = (y - p1y) * (p2x - p1x) / (p2y - p1y) + p1x
+                    if p1x == p2x or x <= xinters:
+                        inside = not inside
+        p1x, p1y = p2x, p2y
+    
+    return inside
+
+def draw_zone(frame, zone_polygon, color=(255, 0, 0), thickness=3):
+    """Draw tracking zone on frame"""
+    pts = np.array(zone_polygon, np.int32)
+    pts = pts.reshape((-1, 1, 2))
+    cv2.polylines(frame, [pts], True, color, thickness)
+    
+    # Add semi-transparent fill
+    overlay = frame.copy()
+    cv2.fillPoly(overlay, [pts], color)
+    cv2.addWeighted(overlay, 0.2, frame, 0.8, 0, frame)
+
+def setup_video_writer(output_path, fps, width, height):
+    """Setup video writer for output"""
+    fourcc = cv2.VideoWriter_fourcc(*'mp4v')
+    return cv2.VideoWriter(output_path, fourcc, fps, (width, height))
+
+def write_frame_to_video(video_writer, frame, repeat_count):
+    """Write frame to video with specified repeat count"""
+    for _ in range(repeat_count):
+        video_writer.write(frame)
+
+def finalize_video(video_writer):
+    """Release video writer"""
+    video_writer.release()
+
+def main():
+    video_path = "sample2.mp4"
+    yolo_model = "bangchakv2/yolov8n.pt"
+    
+    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+    print(f"Using device: {device}")
+    
+    print("Loading YOLO model...")
+    model = YOLO(yolo_model)
+    
+    print("Opening video...")
+    cap = cv2.VideoCapture(video_path)
+    fps = int(cap.get(cv2.CAP_PROP_FPS))
+    width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+    height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+    total_frames = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
+    
+    print(f"Video info: {width}x{height}, {fps} FPS, {total_frames} frames")
+    
+    # Define tracking zone - Gas station floor area (trapezoidal shape)
+    # Based on the perspective of the gas station floor from your image
+    # width 2560, height 1440
+    
+    tracking_zone = [
+        (423, 974),   # Point 1
+        (1540, 1407), # Point 2
+        (1976, 806),  # Point 3
+        (1364, 749)   # Point 4
+    ]
+    
+    print(f"🎯 Tracking zone defined: {tracking_zone}")
+    
+    # CONTINUOUS TRACKING: Process every 118 frames (~2.0s intervals)
+    frame_skip = 118
+    
+    print(f"🎯 CONTINUOUS MODE: Processing every {frame_skip} frames ({frame_skip/fps:.2f}s intervals)")
+    print(f"🎬 Output video will have same duration as input (each processed frame shown for 2 seconds)")
+    print("🔥 ZONE-FIRST TRACKING: Only cars entering the zone will be tracked!")
+    print("Requires 5 consecutive detections IN ZONE for verification")
+    print("🕐 24/7 MODE: Memory reset every hour to prevent overflow")
+    print("Press 'q' to quit")
+    
+    # Setup video writer for output (same fps as input for normal playback speed)
+    output_path = "tracking_output_botsort_zone_track.mp4"
+    output_fps = fps  # Use same fps as input video
+    out = setup_video_writer(output_path, output_fps, width, height)
+    
+    # Track car IDs and their consecutive detections
+    car_id_counts = defaultdict(int)
+    successful_cars = set()
+    last_positions = {}
+    processed_count = 0
+    
+    # ID remapping for clean sequential zone IDs
+    tracker_to_zone_id = {}  # Maps tracker IDs to clean zone IDs
+    next_zone_id = 1  # Next clean zone ID to assign
+    
+    # Store previous frame detections to filter tracking inputs
+    previous_zone_cars = set()
+    
+    # 24/7 operation: Reset every hour (1800 snapshots at 2-sec intervals = 1 hour)
+    RESET_INTERVAL = 1800  # Reset every 1800 processed frames (1 hour)
+    
+    frame_idx = 0
+    
+    while True:
+        # Skip frames to maintain interval
+        for _ in range(frame_skip):
+            ret, frame = cap.read()
+            if not ret:
+                print("\nNo more frames to read")
+                cap.release()
+                cv2.destroyAllWindows()
+                return
+            frame_idx += 1
+        
+        processed_count += 1
+        current_time = frame_idx / fps
+        
+        print(f"\n🎬 Frame {frame_idx} at {current_time:.2f}s (processed #{processed_count})")
+        
+        # 24/7 Memory Management: Reset every hour
+        if processed_count % RESET_INTERVAL == 0:
+            print(f"🕐 HOURLY RESET: Clearing all tracking data (processed {processed_count} frames)")
+            print(f"   📊 Before reset: {len(tracker_to_zone_id)} tracked cars, next Zone ID was {next_zone_id}")
+            
+            # Clear all tracking data
+            tracker_to_zone_id.clear()
+            car_id_counts.clear()
+            successful_cars.clear()
+            last_positions.clear()
+            next_zone_id = 1  # Reset to 1
+            
+            # Reset BoT-SORT tracker state
+            try:
+                model.reset()
+                print(f"   ✅ BoT-SORT tracker reset successfully")
+            except:
+                print(f"   ⚠️ BoT-SORT reset not available (continuing without reset)")
+            
+            print(f"   🆕 Zone IDs will start from 1 again")
+        
+        # Draw tracking zone on frame
+        draw_zone(frame, tracking_zone, color=(0, 255, 255), thickness=3)  # Yellow zone
+        
+        # First run YOLO detection (without tracking) to find cars in zone
+        detection_results = model(frame, verbose=False, conf=0.7, classes=[2])
+        
+        # Find cars currently in the tracking zone
+        current_zone_cars = []
+        total_detections = 0
+        
+        if detection_results[0].boxes is not None:
+            boxes = detection_results[0].boxes.xyxy.cpu()
+            scores = detection_results[0].boxes.conf.cpu()
+            
+            total_detections = len(boxes)
+            print(f"  🔍 Total car detections: {total_detections}")
+            
+            for i in range(len(boxes)):
+                x1, y1, x2, y2 = boxes[i]
+                conf = float(scores[i])
+                
+                # Check if detection is in zone (using bottom center)
+                box_bottom = ((x1 + x2) / 2, y2)
+                if point_in_polygon(box_bottom, tracking_zone):
+                    current_zone_cars.append({
+                        'bbox': [float(x1), float(y1), float(x2), float(y2)],
+                        'conf': conf,
+                        'center': ((x1 + x2) / 2, (y1 + y2) / 2),
+                        'bottom': box_bottom
+                    })
+        
+        print(f"  🎯 Cars in zone: {len(current_zone_cars)}")
+        
+        # Only run tracking if there are cars in the zone
+        detected_car_ids = set()
+        
+        if current_zone_cars:
+            # Run tracking on the full frame (let tracker handle associations)
+            # But we'll filter results to only zone cars afterward
+            results = model.track(
+                frame, 
+                persist=True,
+                verbose=False,
+                conf=0.7,
+                classes=[2],
+                tracker="botsort_reid.yaml"
+            )
+            
+            if results[0].boxes is not None and results[0].boxes.id is not None:
+                boxes = results[0].boxes.xyxy.cpu()
+                scores = results[0].boxes.conf.cpu()
+                track_ids = results[0].boxes.id.cpu().int()
+                
+                print(f"  📊 Total tracked objects: {len(track_ids)}")
+                
+                # Filter tracked objects to only those in zone
+                zone_tracks = []
+                for i, track_id in enumerate(track_ids):
+                    x1, y1, x2, y2 = boxes[i]
+                    conf = float(scores[i])
+                    
+                    # Check if this tracked object is in our zone
+                    box_bottom = ((x1 + x2) / 2, y2)
+                    if point_in_polygon(box_bottom, tracking_zone):
+                        zone_tracks.append({
+                            'id': int(track_id),
+                            'bbox': [int(x1), int(y1), int(x2), int(y2)],
+                            'conf': conf,
+                            'center': ((x1 + x2) / 2, (y1 + y2) / 2),
+                            'bottom': box_bottom
+                        })
+                
+                print(f"  ✅ Zone tracks: {len(zone_tracks)}")
+                
+                # Process each zone track
+                for track in zone_tracks:
+                    tracker_id = track['id']  # Original tracker ID
+                    x1, y1, x2, y2 = track['bbox']
+                    conf = track['conf']
+                    box_center = track['center']
+                    
+                    # Map tracker ID to clean zone ID
+                    if tracker_id not in tracker_to_zone_id:
+                        tracker_to_zone_id[tracker_id] = next_zone_id
+                        print(f"  🆕 New car: Tracker ID {tracker_id} → Zone ID {next_zone_id}")
+                        next_zone_id += 1
+                    
+                    zone_id = tracker_to_zone_id[tracker_id]  # Clean sequential ID
+                    
+                    # Validate track continuity (use tracker_id for internal logic)
+                    is_valid = True
+                    
+                    # Check for suspicious jumps
+                    if tracker_id in last_positions:
+                        last_center = last_positions[tracker_id]
+                        distance = np.sqrt((box_center[0] - last_center[0])**2 + 
+                                         (box_center[1] - last_center[1])**2)
+                        
+                        if distance > 400:  # pixels in ~2.0s
+                            is_valid = False
+                            print(f"  ⚠️ Zone ID {zone_id} (Tracker {tracker_id}): suspicious jump {distance:.0f}px")
+                    
+                    # Skip already successful cars (use zone_id for user logic)
+                    if zone_id in successful_cars:
+                        is_valid = False
+                        print(f"  ✅ Zone ID {zone_id}: already successful, skipping")
+                    
+                    # Only process valid, high-confidence zone tracks
+                    if is_valid and conf > 0.7:
+                        detected_car_ids.add(zone_id)  # Use zone_id for display
+                        car_id_counts[zone_id] += 1
+                        last_positions[tracker_id] = box_center  # Track by tracker_id internally
+                        
+                        # Draw tracking results with clean zone ID
+                        zone_color = (0, 255, 0)  # Green for zone cars
+                        cv2.rectangle(frame, (x1, y1), (x2, y2), zone_color, 2)
+                        cv2.putText(frame, f'ZONE ID:{zone_id}', 
+                                  (x1, y1-30), cv2.FONT_HERSHEY_SIMPLEX, 0.6, zone_color, 2)
+                        cv2.putText(frame, f'#{car_id_counts[zone_id]} {conf:.2f}', 
+                                  (x1, y1-10), cv2.FONT_HERSHEY_SIMPLEX, 0.6, zone_color, 2)
+                        
+                        # Draw center point
+                        cv2.circle(frame, (int(track['bottom'][0]), int(track['bottom'][1])), 5, zone_color, -1)
+                        
+                        print(f"  ✅ Zone ID {zone_id} (Tracker {tracker_id}): ZONE detection #{car_id_counts[zone_id]} (conf: {conf:.2f})")
+                        
+                        # Check for success (5 consecutive detections IN ZONE)
+                        if car_id_counts[zone_id] == 5:
+                            print(f"🏆 SUCCESS: Zone ID {zone_id} achieved 5 continuous ZONE detections - TRIGGER NEXT MODEL!")
+                            successful_cars.add(zone_id)
+                            
+                            # Add success indicator to frame
+                            cv2.putText(frame, f"SUCCESS: Zone Car {zone_id}!", 
+                                      (50, height-50), cv2.FONT_HERSHEY_SIMPLEX, 1.0, (0, 255, 0), 3)
+        else:
+            print("  📋 No cars in zone - no tracking performed")
+            
+            # Draw any cars outside the zone in red (for reference)
+            if detection_results[0].boxes is not None:
+                boxes = detection_results[0].boxes.xyxy.cpu()
+                scores = detection_results[0].boxes.conf.cpu()
+                
+                for i in range(len(boxes)):
+                    x1, y1, x2, y2 = boxes[i]
+                    conf = float(scores[i])
+                    
+                    box_bottom = ((x1 + x2) / 2, y2)
+                    if not point_in_polygon(box_bottom, tracking_zone):
+                        # Draw cars outside zone in red (not tracked)
+                        x1, y1, x2, y2 = int(x1), int(y1), int(x2), int(y2)
+                        cv2.rectangle(frame, (x1, y1), (x2, y2), (0, 0, 255), 1)
+                        cv2.putText(frame, f'OUT {conf:.2f}', 
+                                  (x1, y1-10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 255), 1)
+        
+        # Display results
+        if detected_car_ids:
+            print(f"  📋 Active Zone IDs: {sorted(detected_car_ids)} (Clean sequential IDs)")
+        
+        # Show ID mapping for debugging
+        if tracker_to_zone_id:
+            mapping_str = ", ".join([f"Tracker{k}→Zone{v}" for k, v in tracker_to_zone_id.items()])
+            print(f"  🔄 ID Mapping: {mapping_str}")
+        
+        # Add annotations to frame
+        cv2.putText(frame, f"BoT-SORT Zone-First Tracking | Frame: {frame_idx} | {current_time:.2f}s", 
+                   (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (255, 255, 255), 2)
+        cv2.putText(frame, f"Zone Cars: {len(current_zone_cars)} | Active Tracks: {len(detected_car_ids)}", 
+                   (10, 65), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255, 255, 255), 2)
+        cv2.putText(frame, f"Successful Cars: {len(successful_cars)}", 
+                   (10, 100), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 255, 0), 2)
+        cv2.putText(frame, "TRACKING ZONE", 
+                   (tracking_zone[0][0], tracking_zone[0][1]-10), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 255), 2)
+        
+        # Write annotated frame to output video (repeat for 2 seconds duration)
+        write_frame_to_video(out, frame, frame_skip)
+        
+        # Show video with zone tracking info
+        display_frame = cv2.resize(frame, (960, 540))
+        cv2.imshow('BoT-SORT Zone-First Tracking', display_frame)
+        
+        # Quick check for quit
+        key = cv2.waitKey(1) & 0xFF
+        if key == ord('q'):
+            break
+        
+        # Small delay to see results
+        time.sleep(0.1)
+    
+    cap.release()
+    finalize_video(out)
+    cv2.destroyAllWindows()
+    print(f"\n🎯 BoT-SORT zone-first tracking completed!")
+    print(f"📊 Processed {processed_count} frames with {frame_skip/fps:.2f}s intervals")
+    print(f"🏆 Successfully tracked {len(successful_cars)} unique cars IN ZONE")
+    print(f"💾 Annotated video saved to: {output_path}")
+
+if __name__ == "__main__":
+    main()

view_redis_images.py

@@ -0,0 +1,162 @@
+#!/usr/bin/env python3
+"""
+Script to view frontal images saved in Redis
+"""
+import redis
+import cv2
+import numpy as np
+import sys
+from datetime import datetime
+
+# Redis connection config (from pipeline.json)
+REDIS_CONFIG = {
+    "host": "10.100.1.3",
+    "port": 6379,
+    "password": "FBQgi0i5RevAAMO5Hh66",
+    "db": 0
+}
+
+def connect_redis():
+    """Connect to Redis server."""
+    try:
+        client = redis.Redis(
+            host=REDIS_CONFIG["host"],
+            port=REDIS_CONFIG["port"],
+            password=REDIS_CONFIG["password"],
+            db=REDIS_CONFIG["db"],
+            decode_responses=False  # Keep bytes for images
+        )
+        client.ping()
+        print(f"✅ Connected to Redis at {REDIS_CONFIG['host']}:{REDIS_CONFIG['port']}")
+        return client
+    except redis.exceptions.ConnectionError as e:
+        print(f"❌ Failed to connect to Redis: {e}")
+        return None
+
+def list_image_keys(client):
+    """List all image keys in Redis."""
+    try:
+        # Look for keys matching the inference pattern
+        keys = client.keys("inference:*")
+        print(f"\n📋 Found {len(keys)} image keys:")
+        for i, key in enumerate(keys):
+            key_str = key.decode() if isinstance(key, bytes) else key
+            print(f"{i+1}. {key_str}")
+        return keys
+    except Exception as e:
+        print(f"❌ Error listing keys: {e}")
+        return []
+
+def view_image(client, key):
+    """View a specific image from Redis."""
+    try:
+        # Get image data from Redis
+        image_data = client.get(key)
+        if image_data is None:
+            print(f"❌ No data found for key: {key}")
+            return
+        
+        print(f"📸 Image size: {len(image_data)} bytes")
+        
+        # Convert bytes to numpy array
+        nparr = np.frombuffer(image_data, np.uint8)
+        
+        # Decode image
+        img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
+        if img is None:
+            print("❌ Failed to decode image data")
+            return
+        
+        print(f"🖼️  Image dimensions: {img.shape[1]}x{img.shape[0]} pixels")
+        
+        # Display image
+        key_str = key.decode() if isinstance(key, bytes) else key
+        cv2.imshow(f'Redis Image: {key_str}', img)
+        print("👁️  Image displayed. Press any key to close...")
+        cv2.waitKey(0)
+        cv2.destroyAllWindows()
+        
+        # Ask if user wants to save the image
+        save = input("💾 Save image to file? (y/n): ").lower().strip()
+        if save == 'y':
+            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+            filename = f"redis_image_{timestamp}.jpg"
+            cv2.imwrite(filename, img)
+            print(f"💾 Image saved as: {filename}")
+        
+    except Exception as e:
+        print(f"❌ Error viewing image: {e}")
+
+def monitor_new_images(client):
+    """Monitor for new images being added to Redis."""
+    print("👀 Monitoring for new images... (Press Ctrl+C to stop)")
+    try:
+        # Subscribe to Redis pub/sub for car detections
+        pubsub = client.pubsub()
+        pubsub.subscribe('car_detections')
+        
+        for message in pubsub.listen():
+            if message['type'] == 'message':
+                data = message['data'].decode()
+                print(f"🚨 New detection: {data}")
+                
+                # Try to extract image key from message
+                import json
+                try:
+                    detection_data = json.loads(data)
+                    image_key = detection_data.get('image_key')
+                    if image_key:
+                        print(f"🖼️  New image available: {image_key}")
+                        view_choice = input("View this image now? (y/n): ").lower().strip()
+                        if view_choice == 'y':
+                            view_image(client, image_key)
+                except json.JSONDecodeError:
+                    pass
+                    
+    except KeyboardInterrupt:
+        print("\n👋 Stopping monitor...")
+    except Exception as e:
+        print(f"❌ Monitor error: {e}")
+
+def main():
+    """Main function."""
+    print("🔍 Redis Image Viewer")
+    print("=" * 50)
+    
+    # Connect to Redis
+    client = connect_redis()
+    if not client:
+        return
+    
+    while True:
+        print("\n📋 Options:")
+        print("1. List all image keys")
+        print("2. View specific image")
+        print("3. Monitor for new images")
+        print("4. Exit")
+        
+        choice = input("\nEnter choice (1-4): ").strip()
+        
+        if choice == '1':
+            keys = list_image_keys(client)
+        elif choice == '2':
+            keys = list_image_keys(client)
+            if keys:
+                try:
+                    idx = int(input(f"\nEnter image number (1-{len(keys)}): ")) - 1
+                    if 0 <= idx < len(keys):
+                        view_image(client, keys[idx])
+                    else:
+                        print("❌ Invalid selection")
+                except ValueError:
+                    print("❌ Please enter a valid number")
+        elif choice == '3':
+            monitor_new_images(client)
+        elif choice == '4':
+            print("👋 Goodbye!")
+            break
+        else:
+            print("❌ Invalid choice")
+
+if __name__ == "__main__":
+    main()

webcam_rtsp_server.py

@@ -0,0 +1,325 @@
+#!/usr/bin/env python3
+"""
+Enhanced webcam server that provides both RTSP streaming and HTTP snapshot endpoints
+Compatible with CMS UI requirements for camera configuration
+"""
+
+import cv2
+import threading
+import time
+import logging
+import socket
+from http.server import BaseHTTPRequestHandler, HTTPServer
+import subprocess
+import sys
+import os
+
+# Configure logging
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s: %(message)s"
+)
+logger = logging.getLogger("webcam_rtsp_server")
+
+# Global webcam capture object
+webcam_cap = None
+rtsp_process = None
+
+class WebcamHTTPHandler(BaseHTTPRequestHandler):
+    """HTTP handler for snapshot requests"""
+    
+    def do_GET(self):
+        if self.path == '/snapshot' or self.path == '/snapshot.jpg':
+            try:
+                # Capture fresh frame from webcam for each request
+                ret, frame = webcam_cap.read()
+                if ret and frame is not None:
+                    # Encode as JPEG
+                    success, buffer = cv2.imencode('.jpg', frame, [cv2.IMWRITE_JPEG_QUALITY, 85])
+                    if success:
+                        self.send_response(200)
+                        self.send_header('Content-Type', 'image/jpeg')
+                        self.send_header('Content-Length', str(len(buffer)))
+                        self.send_header('Cache-Control', 'no-cache, no-store, must-revalidate')
+                        self.send_header('Pragma', 'no-cache')
+                        self.send_header('Expires', '0')
+                        self.end_headers()
+                        self.wfile.write(buffer.tobytes())
+                        logger.debug(f"Served webcam snapshot, size: {len(buffer)} bytes")
+                        return
+                    else:
+                        logger.error("Failed to encode frame as JPEG")
+                else:
+                    logger.error("Failed to capture frame from webcam")
+                
+                # Send error response
+                self.send_response(500)
+                self.send_header('Content-Type', 'text/plain')
+                self.end_headers()
+                self.wfile.write(b'Failed to capture webcam frame')
+                
+            except Exception as e:
+                logger.error(f"Error serving snapshot: {e}")
+                self.send_response(500)
+                self.send_header('Content-Type', 'text/plain')
+                self.end_headers()
+                self.wfile.write(f'Error: {str(e)}'.encode())
+        
+        elif self.path == '/status':
+            # Status endpoint for health checking
+            self.send_response(200)
+            self.send_header('Content-Type', 'application/json')
+            self.end_headers()
+            
+            width = int(webcam_cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+            height = int(webcam_cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+            fps = webcam_cap.get(cv2.CAP_PROP_FPS)
+            
+            status = f'{{"status": "online", "width": {width}, "height": {height}, "fps": {fps}}}'
+            self.wfile.write(status.encode())
+        
+        else:
+            # 404 for other paths
+            self.send_response(404)
+            self.send_header('Content-Type', 'text/plain')
+            self.end_headers()
+            self.wfile.write(b'Not Found - Available endpoints: /snapshot, /snapshot.jpg, /status')
+    
+    def log_message(self, format, *args):
+        # Suppress default HTTP server logging to avoid spam
+        pass
+
+def check_ffmpeg():
+    """Check if FFmpeg is available for RTSP streaming"""
+    try:
+        result = subprocess.run(['ffmpeg', '-version'], 
+                              capture_output=True, text=True, timeout=5)
+        if result.returncode == 0:
+            logger.info("FFmpeg found and working")
+            return True
+    except (subprocess.TimeoutExpired, FileNotFoundError, subprocess.SubprocessError):
+        pass
+    
+    logger.warning("FFmpeg not found. RTSP streaming will not be available.")
+    logger.info("To enable RTSP streaming, install FFmpeg:")
+    logger.info("  Windows: Download from https://ffmpeg.org/download.html")
+    logger.info("  Linux: sudo apt install ffmpeg")
+    logger.info("  macOS: brew install ffmpeg")
+    return False
+
+def get_windows_camera_name():
+    """Get the actual camera device name on Windows"""
+    try:
+        # List video devices using FFmpeg with proper encoding handling
+        result = subprocess.run(['ffmpeg', '-f', 'dshow', '-list_devices', 'true', '-i', 'dummy'], 
+                              capture_output=True, text=True, timeout=10, encoding='utf-8', errors='ignore')
+        output = result.stderr  # FFmpeg outputs device list to stderr
+        
+        # Look for video devices in the output
+        lines = output.split('\n')
+        video_devices = []
+        
+        # Parse the output - look for lines with (video) that contain device names in quotes
+        for line in lines:
+            if '[dshow @' in line and '(video)' in line and '"' in line:
+                # Extract device name between first pair of quotes
+                start = line.find('"') + 1
+                end = line.find('"', start)
+                if start > 0 and end > start:
+                    device_name = line[start:end]
+                    video_devices.append(device_name)
+        
+        logger.info(f"Found Windows video devices: {video_devices}")
+        if video_devices:
+            # Force use the first device (index 0) which is the Logitech HD webcam
+            return video_devices[0]  # This will be "罗技高清网络摄像机 C930c"
+        else:
+            logger.info("No devices found via FFmpeg detection, using fallback")
+            # Fall through to fallback names
+        
+    except Exception as e:
+        logger.debug(f"Failed to get Windows camera name: {e}")
+    
+    # Try common camera device names as fallback
+    # Prioritize Integrated Camera since that's what's working now
+    common_names = [
+        "Integrated Camera",  # This is working for the current setup
+        "USB Video Device",  # Common name for USB cameras
+        "USB2.0 Camera",
+        "C930c",  # Direct model name
+        "HD Pro Webcam C930c",  # Full Logitech name
+        "Logitech",  # Brand name
+        "USB Camera",
+        "Webcam"
+    ]
+    logger.info(f"Using fallback camera names: {common_names}")
+    return common_names[0]  # Return "Integrated Camera" first
+
+def start_rtsp_stream(webcam_index=0, rtsp_port=8554):
+    """Start RTSP streaming using FFmpeg"""
+    global rtsp_process
+    
+    if not check_ffmpeg():
+        return None
+    
+    try:
+        # Get the actual camera device name for Windows
+        if sys.platform.startswith('win'):
+            camera_name = get_windows_camera_name()
+            logger.info(f"Using Windows camera device: {camera_name}")
+        
+        # FFmpeg command to stream webcam via RTSP
+        if sys.platform.startswith('win'):
+            cmd = [
+                'ffmpeg',
+                '-f', 'dshow',
+                '-i', f'video={camera_name}',  # Use detected camera name
+                '-c:v', 'libx264',
+                '-preset', 'veryfast',
+                '-tune', 'zerolatency',
+                '-r', '30',
+                '-s', '1280x720',
+                '-f', 'rtsp',
+                f'rtsp://localhost:{rtsp_port}/stream'
+            ]
+        elif sys.platform.startswith('linux'):
+            cmd = [
+                'ffmpeg',
+                '-f', 'v4l2',
+                '-i', f'/dev/video{webcam_index}',
+                '-c:v', 'libx264',
+                '-preset', 'veryfast',
+                '-tune', 'zerolatency',
+                '-r', '30',
+                '-s', '1280x720',
+                '-f', 'rtsp',
+                f'rtsp://localhost:{rtsp_port}/stream'
+            ]
+        else:  # macOS
+            cmd = [
+                'ffmpeg',
+                '-f', 'avfoundation',
+                '-i', f'{webcam_index}:',
+                '-c:v', 'libx264',
+                '-preset', 'veryfast',
+                '-tune', 'zerolatency',
+                '-r', '30',
+                '-s', '1280x720',
+                '-f', 'rtsp',
+                f'rtsp://localhost:{rtsp_port}/stream'
+            ]
+        
+        logger.info(f"Starting RTSP stream on rtsp://localhost:{rtsp_port}/stream")
+        logger.info(f"FFmpeg command: {' '.join(cmd)}")
+        
+        rtsp_process = subprocess.Popen(
+            cmd,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+            text=True
+        )
+        
+        # Give FFmpeg a moment to start
+        time.sleep(2)
+        
+        # Check if process is still running
+        if rtsp_process.poll() is None:
+            logger.info("RTSP streaming started successfully")
+            return rtsp_process
+        else:
+            # Get error output if process failed
+            stdout, stderr = rtsp_process.communicate(timeout=2)
+            logger.error("RTSP streaming failed to start")
+            logger.error(f"FFmpeg stdout: {stdout}")
+            logger.error(f"FFmpeg stderr: {stderr}")
+            return None
+            
+    except Exception as e:
+        logger.error(f"Failed to start RTSP stream: {e}")
+        return None
+
+def get_local_ip():
+    """Get the Wireguard IP address for external access"""
+    # Use Wireguard IP for external access
+    return "10.101.1.4"
+
+def main():
+    global webcam_cap, rtsp_process
+    
+    # Configuration - Force use index 0 for Logitech HD webcam
+    webcam_index = 0  # Logitech HD webcam C930c (1920x1080@30fps)
+    http_port = 8080
+    rtsp_port = 8554
+    
+    logger.info("=== Webcam RTSP & HTTP Server ===")
+    
+    # Initialize webcam
+    logger.info("Initializing webcam...")
+    webcam_cap = cv2.VideoCapture(webcam_index)
+    
+    if not webcam_cap.isOpened():
+        logger.error(f"Failed to open webcam at index {webcam_index}")
+        logger.info("Try different webcam indices (0, 1, 2, etc.)")
+        return
+    
+    # Set webcam properties - Use high resolution for Logitech HD webcam
+    webcam_cap.set(cv2.CAP_PROP_FRAME_WIDTH, 1920)
+    webcam_cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 1080)
+    webcam_cap.set(cv2.CAP_PROP_FPS, 30)
+    
+    width = int(webcam_cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+    height = int(webcam_cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+    fps = webcam_cap.get(cv2.CAP_PROP_FPS)
+    
+    logger.info(f"Webcam initialized: {width}x{height} @ {fps}fps")
+    
+    # Get local IP for CMS configuration
+    local_ip = get_local_ip()
+    
+    # Start RTSP streaming (optional, requires FFmpeg)
+    rtsp_process = start_rtsp_stream(webcam_index, rtsp_port)
+    
+    # Start HTTP server for snapshots
+    server_address = ('0.0.0.0', http_port)  # Bind to all interfaces
+    http_server = HTTPServer(server_address, WebcamHTTPHandler)
+    
+    logger.info("\n=== Server URLs for CMS Configuration ===")
+    logger.info(f"HTTP Snapshot URL: http://{local_ip}:{http_port}/snapshot")
+    
+    if rtsp_process:
+        logger.info(f"RTSP Stream URL: rtsp://{local_ip}:{rtsp_port}/stream")
+    else:
+        logger.info("RTSP Stream: Not available (FFmpeg not found)")
+        logger.info("HTTP-only mode: Use Snapshot URL for camera input")
+    
+    logger.info(f"Status URL: http://{local_ip}:{http_port}/status")
+    logger.info("\n=== CMS Configuration Suggestions ===")
+    logger.info(f"Camera Identifier: webcam-local-01")
+    logger.info(f"RTSP Stream URL: rtsp://{local_ip}:{rtsp_port}/stream")
+    logger.info(f"Snapshot URL: http://{local_ip}:{http_port}/snapshot")
+    logger.info(f"Snapshot Interval: 2000 (ms)")
+    logger.info("\nPress Ctrl+C to stop all servers")
+    
+    try:
+        # Start HTTP server
+        http_server.serve_forever()
+    except KeyboardInterrupt:
+        logger.info("Shutting down servers...")
+    finally:
+        # Clean up
+        if webcam_cap:
+            webcam_cap.release()
+        
+        if rtsp_process:
+            logger.info("Stopping RTSP stream...")
+            rtsp_process.terminate()
+            try:
+                rtsp_process.wait(timeout=5)
+            except subprocess.TimeoutExpired:
+                rtsp_process.kill()
+        
+        http_server.server_close()
+        logger.info("All servers stopped")
+
+if __name__ == "__main__":
+    main()