Implement video processing module with FFmpeg conversion, OpenCV metadata extraction, and file system repository

- Added FFmpegVideoConverter for video format conversion using FFmpeg.
- Implemented NoOpVideoConverter for scenarios where FFmpeg is unavailable.
- Created OpenCVMetadataExtractor for extracting video metadata.
- Developed FileSystemVideoRepository for managing video files in the file system.
- Integrated video services with dependency injection in VideoModule.
- Established API routes for video management and streaming.
- Added request/response schemas for video metadata and streaming information.
- Implemented caching mechanisms for video streaming.
- Included error handling and logging throughout the module.

usda_vision_system/camera/manager.py

@@ -223,7 +223,9 @@ class CameraManager:
 
             # Generate filename with Atlanta timezone timestamp
             timestamp = format_filename_timestamp()
-            filename = f"{camera_name}_recording_{timestamp}.avi"
+            camera_config = self.config.get_camera_by_name(camera_name)
+            video_format = camera_config.video_format if camera_config else "mp4"
+            filename = f"{camera_name}_recording_{timestamp}.{video_format}"
 
             # Start recording
             success = recorder.start_recording(filename)
@@ -283,11 +285,14 @@ class CameraManager:
 
         # Generate filename with datetime prefix
         timestamp = format_filename_timestamp()
+        camera_config = self.config.get_camera_by_name(camera_name)
+        video_format = camera_config.video_format if camera_config else "mp4"
+
         if filename:
             # Always prepend datetime to the provided filename
             filename = f"{timestamp}_{filename}"
         else:
-            filename = f"{camera_name}_manual_{timestamp}.avi"
+            filename = f"{camera_name}_manual_{timestamp}.{video_format}"
 
         return recorder.start_recording(filename)