feat: apply work-plane filter at all three cloud sites

points_via_sam21, _points_via_bbox, and _object_cloud_cv (bbox branch)
now call apply_below_plane_filter on the CV-frame cloud before the
<10-points guard. Responses carry plane_applied + n_points_before_plane.
Without a saved plane the filter is a no-op (plane_applied=false).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src/presa_mano/perception/sam21_bridge.py

@@ -17,6 +17,7 @@ from ..clients import NuwaError, Sam21Error, Sam21Unavailable
 from ..timing import Stopwatch
 from .depth_align import align_depth_to_rgb
 from .depth_to_rgb import CameraRegistration
+from .work_plane import apply_below_plane_filter, get_work_plane
 
 log = logging.getLogger(__name__)
 
@@ -31,13 +32,17 @@ SAM21_BUG_CODES = frozenset({"SHAPE_MISMATCH", "BBOX_OUT_OF_BOUNDS",
 
 
 def points_via_sam21(nuwa, sam21, bbox: tuple[int, int, int, int],
-                     sw: Stopwatch) -> np.ndarray:
+                     sw: Stopwatch) -> tuple[np.ndarray, dict]:
     """SAM21 path: bbox → segmented point cloud in RGB-camera/OpenCV frame.
 
     Returns points in **meters**. Raises Sam21Fallback on failures that
     should trigger the bbox path. Re-raises HTTPException(502) for Nuwa
     failures (the bbox path would fail too — better to fail fast). All
     SAM21-specific logging happens here; callers do not log on fallback.
+
+    After the SAM21 segmentation the work-plane filter is applied (no-op if no
+    plane is calibrated). Returns (points, plane_info) where plane_info has
+    `plane_applied` and `n_points_before_plane`.
     """
     try:
         with sw.step("nuwa_grab"):
@@ -120,7 +125,12 @@ def points_via_sam21(nuwa, sam21, bbox: tuple[int, int, int, int],
             log.error("Unexpected SAM21 error: %r", e)
             raise Sam21Fallback(f"unexpected: {e!r}") from e
 
+    pts_cv, plane_info = apply_below_plane_filter(pts_cv, get_work_plane())
     if pts_cv.shape[0] < 10:
-        log.warning("SAM21 returned %d points (<10), falling back to bbox", pts_cv.shape[0])
-        raise Sam21Fallback(f"too few points: {pts_cv.shape[0]}")
-    return pts_cv
+        log.warning("SAM21 cloud has %d points after plane filter (<10), falling back to bbox",
+                    pts_cv.shape[0])
+        raise Sam21Fallback(
+            f"too few points after plane filter: {pts_cv.shape[0]}"
+            f" (before plane: {plane_info['n_points_before_plane']})"
+        )
+    return pts_cv, plane_info