Setup & Installation
openclaw skills install @roamerxv/ocr-python
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
npx clawhub install ocr-python
What This Skill Does
OCR tool built on PaddleOCR that extracts text from images and PDFs. Supports Chinese and English, including mixed-language documents. Returns recognized text alongside per-line confidence scores.
PaddleOCR handles Chinese-English mixed content reliably, which most OCR libraries do not support without extra configuration.
When to use it
- Extract text from scanned invoice PDFs
- Recognize printed text in product photos
- Convert scanned contracts to editable text
- Pull text from screenshots for data pipelines
- Digitize multi-page printed forms
Example Workflow
Here's how your AI assistant might use this skill in practice.
INPUT
User asks: Extract all text from a scanned multi-page invoice PDF
AGENT
- 1Load the PDF using PyMuPDF and render each page as an image
- 2Initialize PaddleOCR with Chinese-English language support
- 3Run OCR prediction on each page image
- 4Collect rec_texts and rec_scores from each page result
- 5Return combined text with confidence scores per line
OUTPUT
JSON list of recognized text strings with confidence scores for each detected line