Skip to content

Image OCR Module Overview

Purpose

The Image OCR module (md_generator.image) converts Image files or directories into OCR Markdown and metadata. It exists so teams can publish searchable, diff-friendly Markdown from operational inputs without maintaining separate documentation pipelines per format.

Problem solved

  • Manual copy/paste from Image files or directories into wikis does not scale.
  • Downstream AI/RAG workflows need stable text and asset bundles.
  • CI and gateways need a consistent CLI/API surface across mdengine modules.

When to use

  • You need repeatable Markdown export from Image files or directories.
  • You want CLI automation, HTTP conversion, or MCP tool integration (where implemented).
  • You can install the image or image-ocr optional extra (and api for HTTP services).

When not to use

  • Inputs fall outside supported formats or size limits enforced by the module.
  • You require authenticated multi-tenant SaaS features (not provided by this module; use gateway auth).
  • You need transactional writes into application databases (this module documents or transforms; it does not own app schemas except metadata export modules).

Package facts

Item Value
Import path md_generator.image
Source tree src/md_generator/image
CLI md-image
PyPI extra image or image-ocr
Complexity tier medium
API service name image-to-md

Primary entry points

  • convert_images
  • convert_images_recursive
  • build_backends
flowchart LR
    Input[Input] --> Validate[Validate]
    Validate --> Core[image_core]
    Core --> Markdown[Markdown_output]
    Core --> Assets[Optional_assets]