feat(ocr): Add docTR OCR engine with metrics infrastructure

Add docTR as primary OCR engine with 2-tier sequential processing,
OCR metrics tracking, and simplified engine selection.

Features:
- docTR OCR engine with light+medium preprocessing tiers
- doctr_plus mode with early exit optimization (~65% fast path)
- OCR metrics dashboard with per-engine statistics
- User OCR preference persistence
- Parallel worker pool for OCR processing
- Cross-validation for extraction quality

Engine options: tesseract, doctr, doctr_plus (recommended), paddleocr

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

backend/.env.example

@@ -112,6 +112,40 @@ DATA_ENTRY_SQLITE_DATABASE_PATH=data/receipts/receipts.db
 DATA_ENTRY_UPLOAD_PATH=data/receipts/uploads
 DATA_ENTRY_MAX_UPLOAD_SIZE_MB=10
 
+# ============================================================================
+# OCR ENGINE CONFIGURATION
+# ============================================================================
+# Control which OCR engines are loaded at startup.
+# Disabling engines saves memory but limits available OCR modes.
+
+# Enable/disable PaddleOCR (set to 'false' to save ~800MB RAM)
+# When disabled: 'paddleocr' engine unavailable
+OCR_ENABLE_PADDLEOCR=true
+
+# Enable/disable Tesseract (set to 'false' to save ~50MB RAM)
+# When disabled: 'tesseract' engine unavailable
+OCR_ENABLE_TESSERACT=true
+
+# Default OCR engine when not specified in request
+# Options: tesseract, doctr, doctr_plus, paddleocr
+# Recommended: doctr_plus (2-tier sequential with early exit, ~7.5s avg)
+OCR_DEFAULT_ENGINE=doctr_plus
+
+# Active OCR engines shown in frontend dropdown (comma-separated)
+# Options: tesseract, doctr, doctr_plus, paddleocr
+# doctr_plus: 73.3% perfect, 7.5s avg, 65% fast path (recommended)
+# doctr: 63.3% perfect, simpler but faster
+OCR_ACTIVE_ENGINES=tesseract,doctr,doctr_plus,paddleocr
+
+# OCR Worker Pool Configuration
+# Number of parallel OCR workers (each loads ~1GB for docTR)
+# Recommended: 2 for 8GB RAM, 3 for 16GB RAM
+OCR_WORKERS=2
+
+# Max tasks per worker before restart (0 = no restart, saves 40-60s warmup time)
+# Set to 0 for testing, 10-20 for production (prevents memory leaks)
+OCR_MAX_TASKS_PER_CHILD=0
+
 # ============================================================================
 # TELEGRAM MODULE - BOT CONFIGURATION (REQUIRED for Telegram features)
 # ============================================================================