Skip to content

Image OCR Parameter Reference

All parameters below are extracted from source where possible. Validate against --help and OpenAPI (/docs) before production use.

CLI parameters

Name Type Required Default Choices Description
input Path required — — Input image file or directory of images
output Path required — — Output .md path, or output directory with --artifact-layout
--artifact-layout flag optional false — Treat OUTPUT as a directory and write document.md inside it
--engines str optional 'tess,paddle,easy' — Comma-separated engines: tess, paddle, easy (default: tess,paddle,easy)
--strategy str optional 'compare' ('compare', 'best') compare: section per engine; best: longest non-empty text (tie: earlier engine wins)
--title str optional 'OCR extraction' — Top-level Markdown heading
--lang str optional 'eng' — Tesseract language (default: eng)
--paddle-lang str optional 'en' — PaddleOCR lang code (default: en)
--paddle-no-angle-cls flag optional false — Disable PaddleOCR angle classifier
--easy-lang str optional 'en' — Comma-separated EasyOCR language codes (default: en)
--tesseract-cmd str optional None — Path to tesseract executable (default: env TESSERACT_CMD or TESSERACT_PATH)
-v flag optional false — Print dependency/runtime warnings to stderr

API routes (HTTP parameters)

Method Path Type Required Default Description
POST /convert/sync HTTP — — FastAPI route
POST /convert/jobs HTTP — — FastAPI route
GET /convert/jobs/{job_id} HTTP — — FastAPI route
GET /convert/jobs/{job_id}/download HTTP — — FastAPI route

Environment variables

Name Type Required Default Choices Description
IMAGE_TO_MD_TEMP_DIR string/int optional — — From src\md_generator\image\api\settings.py
IMAGE_TO_MD_CORS_ORIGINS string/int optional — — From src\md_generator\image\api\settings.py
IMAGE_TO_MD_TESSERACT_CMD string/int optional — — From src\md_generator\image\api\settings.py
TESSERACT_CMD string/int optional — — From src\md_generator\image\api\settings.py
IMAGE_TO_MD_MAX_UPLOAD_MB string/int optional — — Derived env pattern in settings.py
IMAGE_TO_MD_MAX_SYNC_UPLOAD_MB string/int optional — — Derived env pattern in settings.py
IMAGE_TO_MD_JOB_TTL_SECONDS string/int optional — — Derived env pattern in settings.py
IMAGE_TO_MD_PORT string/int optional — — Derived env pattern in settings.py

Config files

No entries detected.

YAML config keys (from packaged defaults)

No entries detected.

Run config dataclass fields

No entries detected.

Options / dataclass fields (sample)

Class Field Type Required Default Description
ConvertOptions engines tuple[str, ...] varies — Field on options/config class
ConvertOptions strategy Literal['compare', 'best'] varies — Field on options/config class
ConvertOptions title str varies — Field on options/config class
ConvertOptions tess_lang str varies — Field on options/config class
ConvertOptions tesseract_cmd str | None varies — Field on options/config class
ConvertOptions paddle_lang str varies — Field on options/config class
ConvertOptions paddle_use_angle_cls bool varies — Field on options/config class
ConvertOptions easy_langs tuple[str, ...] varies — Field on options/config class
ConvertOptions verbose bool varies — Field on options/config class

Validation and edge cases

  • Positional CLI arguments are required unless documented as optional flags.
  • Upload APIs enforce size limits via environment variables (413 on exceed).
  • Missing optional dependencies raise install-time or runtime errors referencing the PyPI extra image or image-ocr.