style: drop dead Candidate fields, simplify resolve_backend, US spelling

- Candidate carries only the fields the arbiter reads (key, label,
  detected_strict, detected_relaxed, features); location/region/confidence were
  vestigial from the removed best_auto_mark max-by-confidence path.
- resolve_backend returns preferred_inpaint_backend() directly (typed Literal)
  instead of an identity ternary.
- colour/normalise/behaviour -> US spelling across code comments and docs.

No behavior change.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

src/remove_ai_watermarks/cli.py

@@ -699,7 +699,7 @@ def cmd_erase(
     """Erase arbitrary region(s) from an image via inpainting.
 
     Universal and position-agnostic: removes any logo / watermark / object inside
-    the boxes you pass, regardless of colour or location. Runs on CPU. Use this
+    the boxes you pass, regardless of color or location. Runs on CPU. Use this
     for marks the dedicated ``visible`` engines (Gemini, Doubao) do not cover.
     """
     from remove_ai_watermarks.region_eraser import erase

src/remove_ai_watermarks/noai/tiling.py

@@ -81,7 +81,7 @@ def feather_weights(width: int, height: int, overlap: int) -> NDArray[Any]:
 
     Separable linear taper over ``overlap`` pixels from every edge (capped at
     half the tile so short tiles still taper symmetrically). Strictly positive
-    everywhere, so the normalised blend is well-defined even at an image corner
+    everywhere, so the normalized blend is well-defined even at an image corner
     that only one tile covers.
     """
     import numpy as np

src/remove_ai_watermarks/region_eraser.py

@@ -2,7 +2,7 @@
 
 Position- and content-agnostic. You supply the rectangle(s); the eraser inpaints
 whatever is inside, so it removes any visible logo / watermark / object regardless
-of colour, style, or location. Localisation is the user's responsibility (pass the
+of color, style, or location. Localization is the user's responsibility (pass the
 box); restoration runs on CPU. This is the universal fallback for marks the
 deterministic per-generator engines (Gemini sparkle, Doubao) do not cover.
 
@@ -151,7 +151,7 @@ def erase_lama(image_bgr: NDArray[Any], mask: NDArray[Any]) -> NDArray[Any]:
     crop_mask = mask[cy0:cy1, cx0:cx1]
     ch, cw = crop.shape[:2]
 
-    # Resize crop + mask to the model size, normalise to [0,1] RGB CHW.
+    # Resize crop + mask to the model size, normalize to [0,1] RGB CHW.
     crop_rs = cv2.resize(crop, (size, size), interpolation=cv2.INTER_AREA)
     mask_rs = cv2.resize(crop_mask, (size, size), interpolation=cv2.INTER_NEAREST)
     img_in = cv2.cvtColor(crop_rs, cv2.COLOR_BGR2RGB).astype(np.float32) / 255.0

src/remove_ai_watermarks/watermark_registry.py

@@ -11,7 +11,7 @@ that mask to ONE shared, swappable fill backend (``region_eraser``: cv2 Telea/NS
 MI-GAN, or big-LaMa). No mark carries a reverse-alpha step any more: the old
 ``original = (wm - a*logo)/(1-a)`` recovery depended on a fixed captured alpha map
 at a fixed position, broke whenever a vendor re-rendered or moved its mark, and was
-not colour-lossless even with the right map (it amplifies quantization/JPEG-chroma
+not color-lossless even with the right map (it amplifies quantization/JPEG-chroma
 error by ``1/(1-a)`` -- the "the color just changed, not removed" reports). The
 localizer stays cheap (cv2/numpy, CPU) so a memory-tight caller can run it on a
 small worker; the heavy fill (MI-GAN / LaMa) is opt-in and chosen by the caller.
@@ -123,11 +123,8 @@ class Candidate:
 
     key: str
     label: str
-    location: str
-    region: Region
     detected_strict: bool
     detected_relaxed: bool
-    confidence: float
     features: dict[str, float]  # generic; both construction sites always supply it (empty when none)
 
 
@@ -279,7 +276,7 @@ def inpaint_model_available() -> bool:
     return region_eraser.migan_available() or region_eraser.lama_available()
 
 
-def preferred_inpaint_backend() -> str:
+def preferred_inpaint_backend() -> Literal["migan", "cv2"]:
     """Backend used by the ``auto`` fill: MI-GAN (light, droplet-friendly, the
     default) when its ONNX runtime is available, else cv2. big-LaMa is NOT auto-
     selected -- it is a heavier explicit opt-in via ``--backend lama`` (both models
@@ -293,7 +290,7 @@ def preferred_inpaint_backend() -> str:
 def resolve_backend(backend: Backend) -> Literal["cv2", "migan", "lama"]:
     """Resolve ``auto`` to the preferred installed backend; pass the rest through."""
     if backend == "auto":
-        return "migan" if preferred_inpaint_backend() == "migan" else "cv2"
+        return preferred_inpaint_backend()
     return backend
 
 
@@ -494,11 +491,7 @@ def _build_candidates(image: NDArray[Any]) -> list[Candidate]:
         strict = m.detect(image, provenance=False)
         relaxed = m.detect(image, provenance=True)
         feats = m.features(image) if (strict.detected or relaxed.detected) else {}
-        cands.append(
-            Candidate(
-                m.key, m.label, m.location, strict.region, strict.detected, relaxed.detected, strict.confidence, feats
-            )
-        )
+        cands.append(Candidate(m.key, m.label, strict.detected, relaxed.detected, feats))
     return cands