Update ocr_engine.py

ocr_engine.py

@@ -5,54 +5,64 @@ import re
 
 reader = easyocr.Reader(['en'], gpu=False)
 
-def enhance_image(img):
-    # Resize large image down to avoid OCR failure
-    max_dim = 1000
-    height, width = img.shape[:2]
-    if max(height, width) > max_dim:
-        scale = max_dim / max(height, width)
-        img = cv2.resize(img, None, fx=scale, fy=scale, interpolation=cv2.INTER_AREA)
-
-    # Convert to gray
-    gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
-
-    # Denoise
-    gray = cv2.fastNlMeansDenoising(gray, h=15)
-
-    # Sharpen
-    kernel = np.array([[0, -1, 0], [-1, 5,-1], [0, -1, 0]])
-    sharp = cv2.filter2D(gray, -1, kernel)
-
-    # Contrast
-    clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8,8))
-    contrast = clahe.apply(sharp)
-
-    return contrast
-
 def extract_weight_from_image(pil_img):
     try:
         img = np.array(pil_img)
-        enhanced = enhance_image(img)
 
-        results = reader.readtext(enhanced)
+        # Resize if image is too big
+        max_dim = 1000
+        height, width = img.shape[:2]
+        if max(height, width) > max_dim:
+            scale = max_dim / max(height, width)
+            img = cv2.resize(img, None, fx=scale, fy=scale, interpolation=cv2.INTER_AREA)
+
+        # OCR
+        results = reader.readtext(img)
         print("DEBUG OCR RESULTS:", results)
 
-        ocr_texts = [text for _, text, _ in results]
+        raw_texts = []
         weight_candidates = []
 
+        fallback_weight = None
+        fallback_conf = 0.0
+
         for _, text, conf in results:
-            cleaned = text.lower().replace("kg", "").replace("kgs", "")
-            cleaned = cleaned.replace("o", "0").replace("s", "5").replace("g", "9")
-            cleaned = re.sub(r"[^\d\.]", "", cleaned)
+            original = text
+            cleaned = text.lower().strip()
+
+            cleaned = cleaned.replace(",", ".")
+            cleaned = cleaned.replace("o", "0").replace("O", "0")
+            cleaned = cleaned.replace("s", "5").replace("S", "5")
+            cleaned = cleaned.replace("g", "9").replace("G", "6")
+            cleaned = cleaned.replace("kg", "").replace("kgs", "")
+            cleaned = re.sub(r"[^0-9\.]", "", cleaned)
 
-            if re.fullmatch(r"\d{2,4}(\.\d{1,2})?", cleaned):
+            raw_texts.append(f"{original} → {cleaned} (conf: {round(conf, 2)})")
+
+            # Save fallback if any number
+            if cleaned and cleaned.replace(".", "").isdigit() and not fallback_weight:
+                fallback_weight = cleaned
+                fallback_conf = conf
+
+            if cleaned.count(".") <= 1 and re.fullmatch(r"\d{2,4}(\.\d{1,3})?", cleaned):
                 weight_candidates.append((cleaned, conf))
 
-        if not weight_candidates:
-            return "Not detected", 0.0, "\n".join(ocr_texts)
+        if weight_candidates:
+            best_weight, best_conf = sorted(weight_candidates, key=lambda x: -x[1])[0]
+        elif fallback_weight:
+            best_weight, best_conf = fallback_weight, fallback_conf
+        else:
+            return "Not detected", 0.0, "\n".join(raw_texts)
+
+        # Clean up leading zeros
+        if "." in best_weight:
+            int_part, dec_part = best_weight.split(".")
+            int_part = int_part.lstrip("0") or "0"
+            best_weight = f"{int_part}.{dec_part}"
+        else:
+            best_weight = best_weight.lstrip("0") or "0"
 
-        best_weight, best_conf = sorted(weight_candidates, key=lambda x: -x[1])[0]
-        return best_weight, round(best_conf * 100, 2), "\n".join(ocr_texts)
+        return best_weight, round(best_conf * 100, 2), "\n".join(raw_texts)
 
     except Exception as e:
         return f"Error: {str(e)}", 0.0, "OCR failed"