Image Recognition

Extracts text, tables, and visual data from images using PaddleOCR-VL.

2|Updated Mar 10, 2026

Search Tags:#ocr #text extraction #document analysis #image recognition #multilingual #paddleocr

Ready to run

npx skills add https://github.com/caoronglin/copaw-skills --skill image-recognition

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: Image Recognition
Source: https://github.com/caoronglin/copaw-skills/tree/main/skills/image-recognition
Command: npx skills add https://github.com/caoronglin/copaw-skills --skill image-recognition

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, paddleocr, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill tackles the challenge of extracting meaningful information from images, whether it's text, scenes, or objects, making visual content accessible and processable.

Core Features & Use Cases

Text Recognition (OCR): Extracts text from images across 109 languages.
Document Analysis: Identifies text, tables, and formulas within document images.
Table Data Extraction: Pulls structured data from tables in images.
Use Case: Automatically digitize scanned invoices by extracting all text and table data, then convert it into a structured format for easy analysis.

Quick Start

Use the image recognition skill to extract all text from the attached file 'receipt.png'.

Frequently Asked Questions about Image Recognition

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I extract text from images containing multiple languages?▼

To extract text from images containing multiple languages, this Skill uses the PaddleOCR-VL model to recognize and digitize visual text across 109 languages. It processes the input image and returns the recognized text accurately.

What is the best way to digitize scanned invoices and extract table data?▼

The best way to digitize scanned invoices and extract table data is using document analysis features. This Skill identifies text, tables, and formulas within images, pulling structured data and converting it into accessible JSON or Markdown formats.

Does PaddleOCR support document layout analysis and structured JSON output?▼

Yes, PaddleOCR supports document layout analysis and structured JSON output. It identifies text, tables, and formulas within document images, enabling you to extract structured data and output it directly as JSON or Markdown.

Can I use this image recognition tool to extract formulas from document images?▼

Yes, you can use this image recognition tool to extract formulas from document images. The document analysis feature specifically identifies text, tables, and formulas, making visual content fully accessible and processable.

How do I pull structured data from tables in images using OCR?▼

To pull structured data from tables in images using OCR, the Skill performs table data extraction using the PaddleOCR-VL model. It identifies table layouts within the image and extracts the data into structured JSON or Markdown output.

What limitations exist when processing multilingual document images for text extraction?▼

Text extraction supports 109 languages and outputs JSON or Markdown, but processing heavily relies on the PaddleOCR model. Complex visual content or extremely low-quality images may impact the accuracy of layout analysis and structured data extraction.