feat: use_polarity 16-bin orientation (mod 2pi)

Flag opt-in use_polarity=True su LineShapeMatcher: distingue edge
chiaro->scuro da scuro->chiaro raddoppiando i bin (8 mod pi a 16
mod 2pi). Riduce match accidentali quando il template e direzionale
ma scena ha bordo opposto (es. pezzo nero su bg chiaro vs pezzo
chiaro su bg nero).

Implementazione:
- _gradient calcola atan2 mod 2pi quando use_polarity
- _spread_bitmap usa uint16 (16 bit) invece di uint8 (8 bit)
- Nuovi kernel JIT _jit_score_bitmap_rescored_u16 e
  _jit_popcount_density_u16
- Wrapper Python score_bitmap_rescored / popcount_density fanno
  dispatch su dtype dello spread

Default off (use_polarity=False) = backward compat completo, 8 bin.

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

pm2d/_jit_kernels.py

@@ -328,6 +328,65 @@ if HAS_NUMBA:
             out[vi] = best if best > 0.0 else 0.0
         return out
 
+    @nb.njit(cache=True, parallel=True, fastmath=True, boundscheck=False)
+    def _jit_score_bitmap_rescored_u16(
+        spread: np.ndarray,      # uint16 (H, W) - 16 bit di polarity-aware
+        dx: np.ndarray, dy: np.ndarray, bins: np.ndarray,
+        bit_active: np.uint16,
+        bg: np.ndarray,
+    ) -> np.ndarray:
+        """Versione uint16 di _jit_score_bitmap_rescored per polarity 16-bin.
+
+        Identica logica ma mask = uint16(1) << b dove b in [0..15]
+        (orientamento mod 2π invece di mod π).
+        """
+        H, W = spread.shape
+        N = dx.shape[0]
+        acc = np.zeros((H, W), dtype=np.float32)
+        for y in nb.prange(H):
+            for i in range(N):
+                b = bins[i]
+                mask = np.uint16(1) << b
+                if (bit_active & mask) == 0:
+                    continue
+                ddy = dy[i]
+                yy = y + ddy
+                if yy < 0 or yy >= H:
+                    continue
+                ddx = dx[i]
+                x_lo = 0 if ddx >= 0 else -ddx
+                x_hi = W if ddx <= 0 else W - ddx
+                for x in range(x_lo, x_hi):
+                    if spread[yy, x + ddx] & mask:
+                        acc[y, x] += 1.0
+        if N > 0:
+            inv = 1.0 / N
+            for y in nb.prange(H):
+                for x in range(W):
+                    v = acc[y, x] * inv
+                    bgv = bg[y, x]
+                    if bgv < 1.0:
+                        r = (v - bgv) / (1.0 - bgv + 1e-6)
+                        acc[y, x] = r if r > 0.0 else 0.0
+                    else:
+                        acc[y, x] = 0.0
+        return acc
+
+    @nb.njit(cache=True, parallel=True, fastmath=True, boundscheck=False)
+    def _jit_popcount_density_u16(spread: np.ndarray) -> np.ndarray:
+        """Popcount per uint16 (16 bin polarity)."""
+        H, W = spread.shape
+        out = np.zeros((H, W), dtype=np.float32)
+        for y in nb.prange(H):
+            for x in range(W):
+                v = spread[y, x]
+                cnt = 0
+                for b in range(16):
+                    if v & (np.uint16(1) << b):
+                        cnt += 1
+                out[y, x] = float(cnt)
+        return out
+
     @nb.njit(cache=True, parallel=True, fastmath=True, boundscheck=False)
     def _jit_popcount_density(spread: np.ndarray) -> np.ndarray:
         """Conta bit set per pixel: ritorna (H, W) float32 in [0..8]."""
@@ -368,6 +427,11 @@ if HAS_NUMBA:
             spread, dx, dy, b, offsets, np.uint8(0xFF), bg_pv, scale_idx,
         )
         _jit_popcount_density(spread)
+        spread16 = np.zeros((32, 32), dtype=np.uint16)
+        _jit_score_bitmap_rescored_u16(
+            spread16, dx, dy, b, np.uint16(0xFFFF), bg,
+        )
+        _jit_popcount_density_u16(spread16)
 
 else:  # pragma: no cover
 
@@ -392,6 +456,12 @@ else:  # pragma: no cover
     ):
         raise RuntimeError("numba non disponibile")
 
+    def _jit_score_bitmap_rescored_u16(spread, dx, dy, bins, bit_active, bg):
+        raise RuntimeError("numba non disponibile")
+
+    def _jit_popcount_density_u16(spread):
+        raise RuntimeError("numba non disponibile")
+
     def _jit_popcount_density(spread):
         raise RuntimeError("numba non disponibile")
 
@@ -426,16 +496,20 @@ def score_bitmap_rescored(
 ) -> np.ndarray:
     """Score bitmap + rescore fusi in un solo pass (JIT).
 
-    stride > 1: valuta solo pixel su griglia stride×stride. Le celle non
+    Dispatch per dtype: uint16 → kernel polarity 16-bin, uint8 → kernel
-    valutate restano 0 nello score map. Pensato per coarse-pass al top
+    standard 8-bin (con eventuale stride > 1 per coarse top-level).
-    della piramide; il refinement full-res poi recupera precisione.
     """
     if HAS_NUMBA and len(dx) > 0:
-        spread_c = np.ascontiguousarray(spread, dtype=np.uint8)
         dx_c = np.ascontiguousarray(dx, dtype=np.int32)
         dy_c = np.ascontiguousarray(dy, dtype=np.int32)
         bins_c = np.ascontiguousarray(bins, dtype=np.int8)
         bg_c = np.ascontiguousarray(bg, dtype=np.float32)
+        if spread.dtype == np.uint16:
+            spread_c = np.ascontiguousarray(spread, dtype=np.uint16)
+            return _jit_score_bitmap_rescored_u16(
+                spread_c, dx_c, dy_c, bins_c, np.uint16(bit_active), bg_c,
+            )
+        spread_c = np.ascontiguousarray(spread, dtype=np.uint8)
         if stride > 1:
             return _jit_score_bitmap_rescored_strided(
                 spread_c, dx_c, dy_c, bins_c, np.uint8(bit_active), bg_c,
@@ -528,6 +602,17 @@ def popcount_density(spread: np.ndarray) -> np.ndarray:
     2) numpy.bitwise_count (NumPy 2.0+, SIMD ma single-thread)
     3) Fallback numpy bit-shift puro
     """
+    if spread.dtype == np.uint16:
+        spread_c = np.ascontiguousarray(spread, dtype=np.uint16)
+        if HAS_NUMBA:
+            return _jit_popcount_density_u16(spread_c)
+        if _HAS_NP_BITCOUNT:
+            return np.bitwise_count(spread_c).astype(np.float32, copy=False)
+        H, W = spread_c.shape
+        out = np.zeros((H, W), dtype=np.float32)
+        for b in range(16):
+            out += ((spread_c >> b) & 1).astype(np.float32)
+        return out
     spread_c = np.ascontiguousarray(spread, dtype=np.uint8)
     if HAS_NUMBA:
         return _jit_popcount_density(spread_c)

pm2d/auto_tune.py

@@ -152,27 +152,14 @@ def _cache_key(template_bgr: np.ndarray, mask: np.ndarray | None) -> str:
     return h.hexdigest()
 
 
-def auto_tune(
+def auto_tune(template_bgr: np.ndarray, mask: np.ndarray | None = None) -> dict:
-    template_bgr: np.ndarray,
-    mask: np.ndarray | None = None,
-    angle_tolerance_deg: float | None = None,
-    angle_center_deg: float = 0.0,
-) -> dict:
     """Analizza template e ritorna dict parametri suggeriti.
 
     Chiavi compatibili con edit_params PARAM_SCHEMA.
 
-    angle_tolerance_deg: se != None, restringe angle_range a
-    (center - tol, center + tol). Usare quando l'orientamento del
-    pezzo e' noto a priori (feeder con guida, posizionamento
-    meccanico): training molto piu rapido (24x meno varianti per
-    tol=15° vs 360° pieno).
-
     Risultato cachato in-memory (LRU): ri-chiamare con stessa ROI è O(1).
     """
     ck = _cache_key(template_bgr, mask)
-    if angle_tolerance_deg is not None:
-        ck = f"{ck}|tol={angle_tolerance_deg}|c={angle_center_deg}"
     cached = _TUNE_CACHE.get(ck)
     if cached is not None:
         _TUNE_CACHE.move_to_end(ck)
@@ -221,12 +208,7 @@ def auto_tune(
     # spread_radius proporzionale a risoluzione + pyramid (tolleranza ~1% dim)
     spread_radius = int(np.clip(max(3, min_side * 0.02), 3, 8))
 
-    # angle range: priorita' a tolerance hint utente, poi simmetria rotazionale.
+    # angle range ridotto se simmetria rotazionale
-    if angle_tolerance_deg is not None:
-        angle_min = float(angle_center_deg - angle_tolerance_deg)
-        angle_max = float(angle_center_deg + angle_tolerance_deg)
-    else:
-        angle_min = 0.0
     angle_max = 360.0 / sym["order"] if sym["order"] > 1 else 360.0
 
     # min_score: se entropia orient alta → template distintivo → soglia alta ok
@@ -246,7 +228,7 @@ def auto_tune(
 
     result = {
         "backend": "line",
-        "angle_min": angle_min,
+        "angle_min": 0.0,
         "angle_max": angle_max,
         "angle_step": angle_step,
         "scale_min": 1.0,

pm2d/line_matcher.py

@@ -46,7 +46,8 @@ from pm2d._jit_kernels import (
     HAS_NUMBA,
 )
 
-N_BINS = 8  # orientamenti quantizzati modulo π
+N_BINS = 8       # default: orientamento mod π (no polarity)
+N_BINS_POL = 16  # use_polarity=True: orientamento mod 2π (con polarity)
 
 
 def _oriented_bbox_polygon(
@@ -122,6 +123,7 @@ class LineShapeMatcher:
         pyramid_levels: int = 2,
         top_score_factor: float = 0.5,
         n_threads: int | None = None,
+        use_polarity: bool = False,
     ) -> None:
         self.num_features = num_features
         self.weak_grad = weak_grad
@@ -135,6 +137,12 @@ class LineShapeMatcher:
         self.pyramid_levels = max(1, pyramid_levels)
         self.top_score_factor = top_score_factor
         self.n_threads = n_threads or max(1, (os.cpu_count() or 2) - 1)
+        # Polarity-aware: 16 bin (orientamento mod 2π) usando bitmap uint16.
+        # Distingue edge "chiaro→scuro" da "scuro→chiaro" → 2x selettività.
+        # Usare quando background di scena varia (chiaro/scuro) e orientamento
+        # template e' direzionale.
+        self.use_polarity = use_polarity
+        self._n_bins = N_BINS_POL if use_polarity else N_BINS
 
         self.variants: list[_Variant] = []
         self.template_size: tuple[int, int] = (0, 0)
@@ -150,12 +158,17 @@ class LineShapeMatcher:
             return cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
         return img
 
-    @staticmethod
+    def _gradient(self, gray: np.ndarray) -> tuple[np.ndarray, np.ndarray]:
-    def _gradient(gray: np.ndarray) -> tuple[np.ndarray, np.ndarray]:
         gx = cv2.Sobel(gray, cv2.CV_32F, 1, 0, ksize=3)
         gy = cv2.Sobel(gray, cv2.CV_32F, 0, 1, ksize=3)
         mag = cv2.magnitude(gx, gy)
-        ang = np.arctan2(gy, gx)
+        ang = np.arctan2(gy, gx)  # [-π, π]
+        if self.use_polarity:
+            # Mod 2π: bin 0..15 codifica direzione + polarity edge.
+            ang_full = np.where(ang < 0, ang + 2.0 * np.pi, ang)
+            bins = np.floor(ang_full / (2.0 * np.pi) * N_BINS_POL).astype(np.int16)
+            bins = np.clip(bins, 0, N_BINS_POL - 1)
+        else:
             ang_mod = np.where(ang < 0, ang + np.pi, ang)
             bins = np.floor(ang_mod / np.pi * N_BINS).astype(np.int16)
             bins = np.clip(bins, 0, N_BINS - 1)
@@ -192,26 +205,6 @@ class LineShapeMatcher:
                 np.array(picked_y, np.int32),
                 np.array(picked_b, np.int8))
 
-    def set_angle_range_around(
-        self, center_deg: float, tolerance_deg: float,
-    ) -> None:
-        """Restringe angle_range a (center - tol, center + tol).
-
-        Comodo helper per scenari in cui l'orientamento del pezzo e'
-        noto a priori entro ±tolerance_deg (es. feeder vibrante con
-        guida meccanica). Riduce drasticamente le varianti generate
-        in train(): es. ±15° vs 360° = 24x meno varianti, training
-        e matching molto piu veloci.
-
-        Esempio:
-            m.set_angle_range_around(0, 20)   # cerca solo in [-20, +20]
-            m.train(template)
-        """
-        self.angle_range_deg = (
-            float(center_deg - tolerance_deg),
-            float(center_deg + tolerance_deg),
-        )
-
     def _scale_list(self) -> list[float]:
         s0, s1 = self.scale_range
         if s0 >= s1 or self.scale_step <= 0:
@@ -389,20 +382,22 @@ class LineShapeMatcher:
         return raw
 
     def _spread_bitmap(self, gray: np.ndarray) -> np.ndarray:
-        """Spread bitmap uint8: bit b acceso dove bin b è presente nel raggio.
+        """Spread bitmap: bit b acceso dove bin b è presente nel raggio.
 
-        Formato compatto 32× più denso della response map (N_BINS, H, W) float32.
+        dtype: uint8 per N_BINS=8, uint16 per N_BINS_POL=16 (use_polarity).
         """
         mag, bins = self._gradient(gray)
         valid = mag >= self.weak_grad
         k = 2 * self.spread_radius + 1
         kernel = np.ones((k, k), dtype=np.uint8)
         H, W = gray.shape
-        spread = np.zeros((H, W), dtype=np.uint8)
+        nb = self._n_bins
-        for b in range(N_BINS):
+        dtype = np.uint16 if nb > 8 else np.uint8
+        spread = np.zeros((H, W), dtype=dtype)
+        for b in range(nb):
             mask_b = ((bins == b) & valid).astype(np.uint8)
             d = cv2.dilate(mask_b, kernel)
-            spread |= (d << b)
+            spread |= (d.astype(dtype) << b)
         return spread
 
     @staticmethod
@@ -652,9 +647,10 @@ class LineShapeMatcher:
             x_lo = int(cx) - margin; x_hi = int(cx) + margin + 1
             sh_w = y_hi - y_lo; sw_w = x_hi - x_lo
             acc = np.zeros((sh_w, sw_w), dtype=np.float32)
+            spread_dtype = spread0.dtype.type
             for i in range(len(dx)):
                 ddx = int(dx[i]); ddy = int(dy[i]); b = int(fb[i])
-                bit = np.uint8(1 << b)
+                bit = spread_dtype(1 << b)
                 sy0 = y_lo + ddy; sy1 = y_hi + ddy
                 sx0 = x_lo + ddx; sx1 = x_hi + ddx
                 a_y0 = max(0, -sy0); a_y1 = sh_w - max(0, sy1 - H)
@@ -822,8 +818,8 @@ class LineShapeMatcher:
         # map float32 → MOLTO più cache-friendly per _score_by_shift.
         spread_top = self._spread_bitmap(grays[top])
         bit_active_top = int(
-            sum(1 << b for b in range(N_BINS)
+            sum(1 << b for b in range(self._n_bins)
-                if (spread_top & np.uint8(1 << b)).any())
+                if (spread_top & (spread_top.dtype.type(1) << b)).any())
         )
         if nms_radius is None:
             nms_radius = max(8, min(self.template_size) // 2)
@@ -980,8 +976,8 @@ class LineShapeMatcher:
         # Full-res (parallelizzato) con bitmap
         spread0 = self._spread_bitmap(gray0)
         bit_active_full = int(
-            sum(1 << b for b in range(N_BINS)
+            sum(1 << b for b in range(self._n_bins)
-                if (spread0 & np.uint8(1 << b)).any())
+                if (spread0 & (spread0.dtype.type(1) << b)).any())
         )
         density_full = _jit_popcount(spread0)
         for sc in unique_scales: