feat: automatische Orientierungserkennung fuer umgedrehte Scans

Tesseract OSD erkennt 0/90/180/270° Rotation und korrigiert
automatisch vor dem Deskew. Loest das Problem mit Buchscannern,
bei denen jede 2. Seite auf dem Kopf steht.

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

klausur-service/backend/vocab_worksheet_api.py

@@ -71,6 +71,7 @@ try:
         detect_row_geometry, build_cell_grid_v2,
         _cells_to_vocab_entries, _detect_sub_columns, _detect_header_footer_gaps,
         expand_narrow_columns, positional_column_regions, llm_review_entries,
+        detect_and_fix_orientation,
         _fix_phonetic_brackets,
         render_pdf_high_res,
         PageRegion, RowGeometry,
@@ -1360,6 +1361,15 @@ async def _run_ocr_pipeline_for_page(
     img_h, img_w = img_bgr.shape[:2]
     logger.info(f"OCR Pipeline page {page_number + 1}: image {img_w}x{img_h}")
 
+    # 1b. Orientation detection (fix upside-down scans)
+    t0 = _time.time()
+    img_bgr, rotation = detect_and_fix_orientation(img_bgr)
+    if rotation:
+        img_h, img_w = img_bgr.shape[:2]
+        logger.info(f"  orientation: rotated {rotation}° ({_time.time() - t0:.1f}s)")
+    else:
+        logger.info(f"  orientation: OK ({_time.time() - t0:.1f}s)")
+
     # 2. Create pipeline session in DB (for debugging in admin UI)
     pipeline_session_id = str(uuid.uuid4())
     try: