OCR Engines

OCR Engines

OculiX introduces a pluggable OCR architecture with PaddleOCR as primary engine and Tesseract as fallback.

---

Architecture

OCREngine (interface)
  ├── PaddleOCREngine   → HTTP client, zero external deps
  │     └── PaddleOCRClient  → JSON parsing, connection handling (629 lines)
  └── TesseractEngine   → Tess4J wrapper (247 lines)

Auto-detection: PaddleOCR is tried first. If unavailable, falls back to Tesseract.

PaddleOCR Integration

Component	Description
PaddleOCREngine	Engine adapter (629 lines)
PaddleOCRClient	Zero-dependency HTTP client with manual JSON parsing
Protocol	HTTP REST API to a PaddleOCR server

Utilities

Class	Purpose
AmountVariantGenerator	Generates tolerance variants for monetary formats (e.g. 1,234.56 ↔ 1234.56 ↔ 1 234,56)
TextNormalizer	Accent stripping, case-insensitive comparison, whitespace normalization

Key Classes

All in com.sikulix.ocr.*:

File	Lines	Purpose
OCREngine.java	—	Interface for pluggable engines
PaddleOCREngine.java	629	PaddleOCR implementation
PaddleOCRClient.java	629	HTTP client, zero-dep
TesseractEngine.java	247	Tesseract fallback
AmountVariantGenerator.java	93	Monetary format tolerance
TextNormalizer.java	77	Text normalization