update OCR

src/inference.py

@@ -35,13 +35,23 @@ def get_det_model(checkpoint="best.pt"):
         _det_model = RTDETR(checkpoint)
     return _det_model
 
-# Thêm Surya OCR làm engine thứ 3
-from surya.ocr import run_ocr
-from surya.model.detection.model import load_det_processor, load_det_model
-from surya.model.recognition.model import load_rec_model
-from surya.model.recognition.processor import load_rec_processor
+# ============================================================
+# SURYA OCR (optional)
+# ============================================================
+SURYA_AVAILABLE = False
+try:
+    from surya.ocr import run_ocr
+    from surya.model.detection.model import load_det_processor, load_det_model
+    from surya.model.recognition.model import load_rec_model
+    from surya.model.recognition.processor import load_rec_processor
+    SURYA_AVAILABLE = True
+    print("[INFO] Surya OCR is available")
+except ImportError:
+    print("[WARN] Surya OCR not installed. Install with: pip install surya-ocr")
 
 def ocr_with_surya(img_bgr, langs=["vi", "en"]):
+    if not SURYA_AVAILABLE:
+        raise ImportError("Surya OCR is not installed. Please run: pip install surya-ocr")
     det_processor, det_model = load_det_processor(), load_det_model()
     rec_model, rec_processor = load_rec_model(), load_rec_processor()
     from PIL import Image
@@ -256,16 +266,14 @@ def multi_pass_ocr(img_bgr, reader, ocr_type="note"):
 # DUAL-ENGINE OCR — PaddleOCR (vi) + PaddleOCR (en), chọn tốt hơn
 # ============================================================
 def run_ocr_with_backend(img_bgr, backend="paddle", ocr_type="note"):
-    """
-    Chạy OCR với backend được chọn.
-    backend: "paddle", "easyocr", "surya"
-    Trả về (list_of_texts, avg_confidence) - với surya, confidence luôn = 1.0
-    """
     if backend == "surya":
-        text = ocr_with_surya(img_bgr, langs=["vi", "en"])
-        lines = [line.strip() for line in text.split("\n") if line.strip()]
-        return lines, 1.0  # Surya không trả confidence, coi như 1.0
-    
+        if not SURYA_AVAILABLE:
+            print("[WARN] Surya not available, falling back to paddle")
+            backend = "paddle"
+        else:
+            text = ocr_with_surya(img_bgr, langs=["vi", "en"])
+            lines = [line.strip() for line in text.split("\n") if line.strip()]
+            return lines, 1.0
     # logic cũ cho paddle + easyocr
     reader_vi = get_paddle_reader('vi') if backend == "paddle" else None
     reader_en = get_paddle_reader('en') if backend == "paddle" else None