feat(ocr): add image preprocessing module with CLAHE + grayscale + blur

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

ocr-service/preprocessing.py

@@ -0,0 +1,49 @@
+"""Image preprocessing pipeline for aged document OCR quality improvement."""
+
+import logging
+import os
+
+import cv2
+import numpy as np
+from PIL import Image
+
+logger = logging.getLogger(__name__)
+
+CLAHE_CLIP_LIMIT = float(os.environ.get("OCR_CLAHE_CLIP_LIMIT", "2.0"))
+CLAHE_TILE_SIZE = int(os.environ.get("OCR_CLAHE_TILE_SIZE", "8"))
+
+
+def preprocess_page(image: Image.Image) -> Image.Image:
+    """Apply CLAHE + grayscale + Gaussian blur to improve OCR quality on aged documents.
+
+    Falls back silently to the original image if any step fails.
+    """
+    try:
+        img_array = np.array(image)
+        lab = cv2.cvtColor(img_array, cv2.COLOR_RGB2LAB)
+        del img_array
+
+        l_channel = lab[:, :, 0].copy()
+        del lab
+
+        clahe = cv2.createCLAHE(
+            clipLimit=CLAHE_CLIP_LIMIT,
+            tileGridSize=(CLAHE_TILE_SIZE, CLAHE_TILE_SIZE),
+        )
+        l_clahe = clahe.apply(l_channel)
+        del l_channel
+
+        blurred = cv2.GaussianBlur(l_clahe, (3, 3), 0)
+        del l_clahe
+
+        result = Image.fromarray(blurred, mode="L")
+        del blurred
+
+        return result
+    except Exception as e:
+        logger.warning(
+            "preprocess_page failed (falling back to original): %s: %s",
+            type(e).__name__,
+            e,
+        )
+        return image