feat(vision): il server esegue tramite worker e salva misure vere

Aggiunge POST /api/vision/execute: route l'immagine al worker di visione,
mappa le uscite del grafo sulle quote e le salva con save_measurement -
stesso verdetto, stesso gate del fuori tolleranza di ogni altra misura.
Nuova tabella vision_results (una riga per acquisizione, non per quota) e
input_method 'camera' su measurements.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014BBnuACZSCJqXrMYC3LUMU

src/backend/models/api/vision.py

@@ -0,0 +1,16 @@
+"""Pydantic schemas for the vision execution response."""
+from pydantic import BaseModel
+
+
+class VisionMeasurementResult(BaseModel):
+    """One quote's outcome, as much of it as the client needs to show."""
+    subtask_id: int
+    value: float
+    pass_fail: str
+
+
+class VisionExecuteResponse(BaseModel):
+    """What `POST /api/vision/execute` returns: the run, and every quote it filled."""
+    vision_result_id: int
+    engine_version: str
+    measurements: list[VisionMeasurementResult]