refactor(cameras): remove fps and backend specification find-cameras utils

Author: imstevenpmwork

lerobot/common/cameras/intel/configuration_realsense.py

@@ -69,11 +69,3 @@ def __post_init__(self):
             raise ValueError(
                 f"One of them must be set: name or serial_number, but {self.name=} and {self.serial_number=} provided."
             )
-
-        at_least_one_is_not_none = self.fps is not None or self.width is not None or self.height is not None
-        at_least_one_is_none = self.fps is None or self.width is None or self.height is None
-        if at_least_one_is_not_none and at_least_one_is_none:
-            raise ValueError(
-                "For `fps`, `width` and `height`, either all of them need to be set, or none of them, "
-                f"but {self.fps=}, {self.width=}, {self.height=} were provided."
-            )

lerobot/common/cameras/opencv/camera_opencv.py

@@ -59,7 +59,7 @@ class OpenCVCamera(Camera):
     or port changes, especially on Linux. Use the provided utility script to find
     available camera indices or paths:
     ```bash
-    NOTE(Steven): Point to future util
+    python -m lerobot.find_cameras
     ```
 
     The camera's default settings (FPS, resolution, color mode) are used unless
@@ -132,7 +132,7 @@ def __init__(self, config: OpenCVCameraConfig):
         self.logs: dict = {}  # NOTE(Steven): Might be removed in the future
 
         self.rotation: int | None = get_cv2_rotation(config.rotation)
-        self.backend: int = get_cv2_backend()
+        self.backend: int = get_cv2_backend()  # NOTE(Steven): If I specify backend the opencv open fails
 
     def __str__(self) -> str:
         """Returns a string representation of the camera instance."""
@@ -195,7 +195,7 @@ def connect(self):
         cv2.setNumThreads(1)
 
         logger.debug(f"Attempting to connect to camera {self.index_or_path} using backend {self.backend}...")
-        self.videocapture_camera = cv2.VideoCapture(self.index_or_path, self.backend)
+        self.videocapture_camera = cv2.VideoCapture(self.index_or_path)
 
         if not self.videocapture_camera.isOpened():
             self.videocapture_camera.release()
@@ -303,7 +303,7 @@ def find_cameras(
             targets_to_scan = list(range(max_index_search_range))
 
         for target in targets_to_scan:
-            camera = cv2.VideoCapture(target, get_cv2_backend())
+            camera = cv2.VideoCapture(target)
             if camera.isOpened():
                 default_width = int(camera.get(cv2.CAP_PROP_FRAME_WIDTH))
                 default_height = int(camera.get(cv2.CAP_PROP_FRAME_HEIGHT))

lerobot/find_cameras.py

@@ -32,7 +32,7 @@
 from lerobot.common.cameras.opencv.configuration_opencv import OpenCVCameraConfig
 
 logger = logging.getLogger(__name__)
-logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(module)s - %(message)s")
+# logging.basicConfig(level=logging.INFO, format="%(asctime)s - %(levelname)s - %(module)s - %(message)s")
 
 
 def find_all_opencv_cameras() -> List[Dict[str, Any]]:
@@ -160,10 +160,9 @@ def create_camera_instance(cam_meta: Dict[str, Any]) -> Optional[Dict[str, Any]]
     default_profile = cam_meta.get("default_stream_profile")
     width = default_profile.get("width")
     height = default_profile.get("height")
-    fps = default_profile.get("fps")
     instance = None
 
-    logger.info(f"Preparing {cam_type} ID {cam_id} with profile: Width={width}, Height={height}, FPS={fps}")
+    logger.info(f"Preparing {cam_type} ID {cam_id} with profile: Width={width}, Height={height}")
 
     try:
         if cam_type == "OpenCV":
@@ -172,7 +171,6 @@ def create_camera_instance(cam_meta: Dict[str, Any]) -> Optional[Dict[str, Any]]
                 color_mode=ColorMode.RGB,
                 width=width,
                 height=height,
-                fps=fps,
             )
             instance = OpenCVCamera(cv_config)
         elif cam_type == "RealSense":
@@ -181,7 +179,6 @@ def create_camera_instance(cam_meta: Dict[str, Any]) -> Optional[Dict[str, Any]]
                 color_mode=ColorMode.RGB,
                 width=width,
                 height=height,
-                fps=fps,
             )
             instance = RealSenseCamera(rs_config)
         else:
@@ -191,7 +188,6 @@ def create_camera_instance(cam_meta: Dict[str, Any]) -> Optional[Dict[str, Any]]
         if instance:
             logger.info(f"Connecting to {cam_type} camera: {cam_id}...")
             instance.connect()
-            time.sleep(0.1)
             return {"instance": instance, "meta": cam_meta}
     except Exception as e:
         logger.error(f"Failed to connect or configure {cam_type} camera {cam_id}: {e}")
@@ -211,11 +207,6 @@ def process_camera_image(
 
     try:
         image_data = cam.read()
-        if image_data is None:
-            logger.warning(
-                f"No frame received from {cam_type_str} camera {cam_id_str} at time {current_time:.2f}s."
-            )
-            return None
 
         return save_image(
             image_data,
@@ -293,8 +284,6 @@ def save_images_from_all_cameras(
                 if futures:
                     concurrent.futures.wait(futures)
 
-                time.sleep(0.05)
-
         except KeyboardInterrupt:
             logger.info("Capture interrupted by user.")
         finally:
@@ -360,7 +349,6 @@ def save_images_from_all_cameras(
     args = parser.parse_args()
 
     if args.command is None:
-
         default_output_dir = capture_parser.get_default("output_dir")
         default_record_time_s = capture_parser.get_default("record_time_s")
 
@@ -370,4 +358,4 @@ def save_images_from_all_cameras(
             camera_type_filter=None,
         )
     else:
-        args.func(args)
+        args.func(args)