docs(wiki): record fast-alpr as the evaluated ANPR recognizer baseline

Research note from the recognizer-options query. fast-alpr v0.4.0 (MIT) — a swappable
YOLOv9-detector + CCT-OCR pipeline on ONNX Runtime, CPU-only and offline — fits the
decided vision-service architecture and is MIT end-to-end (code + published weights),
so the ANPR path may not need the scoped AGPL exception. Flags the open caveats:
verify model-weight provenance, and benchmark AL-plate accuracy (default global vs.
the 40+ country EU model). fast-alpr is plate-only, so the vehicle-verification job
stays ours to build. Decision kept open. Updates opencv-anpr-service (new "Recognizer
evaluation" section + licensing nuance), vision-service (open/next), index, log.

Claude-Session: https://claude.ai/code/session_01Xcm6ikLgGoCxxHrxtjkk5V

wiki/entities/opencv-anpr-service.md

@@ -50,8 +50,11 @@ recognition **host-side on ordinary IP-camera snapshots**, replacing the dedicat
 ## Licensing — scoped AGPL exception (amends the standing rule)
 
 The app is strictly **MIT/Apache/BSD** ([[technology-stack]], [[standing-decisions]]). Accurate
-ANPR/vehicle models are mostly **AGPL** (YOLO/Ultralytics detectors, OpenALPR) or commercial.
-Decision (2026-06-15): **allow AGPL inside this service only.** It is a **separate process**, not
+ANPR/vehicle models were *assumed* to be mostly **AGPL** (Ultralytics YOLO detectors, OpenALPR) or
+commercial — but the **fast-alpr stack (above) is MIT end-to-end**, so a permissive ANPR baseline now
+looks achievable (pending the weight-provenance caveat). The exception below still matters for the
+*strongest* models (Ultralytics YOLO) and for the vehicle-verification job. Decision (2026-06-15):
+**allow AGPL inside this service only.** It is a **separate process**, not
 linked into the app, so its obligations don't reach the Node/React codebase; the app's permissive
 guarantee is preserved. Recorded as an explicit exception in [[standing-decisions]] /
 [[vision-service]].
@@ -61,6 +64,51 @@ guarantee is preserved. Recorded as an explicit exception in [[standing-decision
   network-use clause could require offering the service's source — relevant only if productised
   beyond the on-site appliance; flag at that point.
 
+## Recognizer evaluation — fast-alpr is the leading baseline (2026-06-19)
+
+`YOLO vs OpenCV` is a **category error** — they're different pipeline layers, not competitors. ANPR
+is a **pipeline**: (1) plate **detection** (find the box → YOLO-family detector), (2) plate **OCR**
+(read the crop → a CRNN/CCT or OCR engine), (3) **glue** (capture/crop/deskew/draw → OpenCV,
+Apache-2.0, always present). So the real choice is *which end-to-end recognizer*, and **OpenCV is
+used regardless** as the image-handling toolkit.
+
+**Leading option: [fast-alpr](https://github.com/ankandrew/fast-alpr) (v0.4.0, 15 Mar 2026).** A thin
+orchestrator over two **swappable** stages, both on **ONNX Runtime** — which matches THIS service's
+decided architecture (separate localhost Python process, offline, swappable behind an interface)
+almost exactly:
+
+| Stage | Default model | Library | License |
+| --- | --- | --- | --- |
+| Plate detection | `yolo-v9-t-384-license-plate-end2end` | [open-image-models](https://github.com/ankandrew/open-image-models) | MIT |
+| Plate OCR | `cct-xs-v2-global-model` | [fast-plate-ocr](https://github.com/ankandrew/fast-plate-ocr) | MIT |
+
+- **MIT top-to-bottom** (library *and* the published model weights), one maintainer (ankandrew) across
+  all three repos. **The detector is open-image-models' own YOLOv9 ONNX export — NOT the Ultralytics
+  AGPL package** — so fast-alpr is a **permissive baseline that may not even need the scoped AGPL
+  exception** below. ⚠️ **Caveat (verify before relying on it):** a repo's LICENSE covers its *code*;
+  redistributed model *weights* can carry separate provenance (YOLOv9 upstream is GPL-3.0; Ultralytics
+  YOLO is AGPL). Confirm the weight training/provenance (model card) before treating "MIT weights" as
+  settled for compliance — the AGPL-in-service exception is the safety net if it doesn't hold.
+- **CPU-only + fully offline.** No runtime ships by default; pick a backend extra — `fast-alpr[onnx]`
+  (CPU), or `[onnx-gpu]`/`[onnx-openvino]`/`[onnx-directml]`/`[onnx-qnn]` — which maps onto the
+  "CPU now, small GPU/NPU later" compute question ([[bom]], [[open-questions]]).
+- **Albanian/EU plates:** fast-plate-ocr also has a **European model trained on 40+ countries** (newer
+  than the default global model) — benchmark it against the default for AL accuracy.
+- **Modular, no lock-in:** swap either stage via `BaseDetector`/`BaseOCR` (their docs plug in
+  Tesseract). So fast-alpr is the baseline you keep while replacing one stage if needed.
+
+**Scope: fast-alpr is plate-only — it does Job 1 (ANPR) but NOT Job 2 (vehicle verification).** The
+anti-spoofing vehicle-attribute/fingerprint stage is still ours to build — but since fast-alpr already
+standardizes on **ONNX Runtime + a YOLO-family detector**, the vehicle stage shares that runtime (the
+coherent outcome). Other options, weaker: **OpenALPR** (permissive but largely unmaintained, the old
+"permissive-only, weaker" path); **Ultralytics YOLO + PaddleOCR** (most accurate/tunable, but YOLO is
+AGPL → needs the in-service exception; most build effort — the "scale" path if fast-alpr's accuracy
+disappoints).
+
+**Recommendation:** prototype with **fast-alpr** now (permissive, offline, ONNX, fits the decided
+shape); plan a YOLO-detector fine-tune + PaddleOCR only if production accuracy demands it. Choice kept
+**open** pending the weight-provenance check + an accuracy benchmark on real AL plates.
+
 ## Anti-fraud / threat-model fit
 
 - **Plate spoofing** (the motivating case): vehicle-attribute / fingerprint mismatch entry↔exit or
@@ -75,8 +123,10 @@ guarantee is preserved. Recorded as an explicit exception in [[standing-decision
 
 ## Open
 
-- **Recognizer choice** (permissive-only vs. AGPL model) and accuracy targets — see
-  [[vision-service]]; AGPL now permitted in-service.
+- **Recognizer choice** — **fast-alpr (MIT, YOLOv9+CCT on ONNX) is the evaluated baseline** (see the
+  Recognizer evaluation section above); remaining open items are the **weight-provenance check** and
+  an **accuracy benchmark on real AL plates** (default global vs. the 40+country EU model). See
+  [[vision-service]]; AGPL still permitted in-service for the stronger fallback.
 - **Vehicle fingerprint**: attribute classifier vs. embedding-similarity; what threshold makes a
   mismatch an anomaly without false-positiving on lighting/angle.
 - **Compute footprint** on the appliance (CPU-only vs. a small GPU/NPU) — procurement input