OCR Node

Extract text from images and documents.

Overview

The OCR (Optical Character Recognition) node extracts text content from images, PDFs, and scanned documents. It converts visual text into machine-readable text that can be processed by other nodes.

Configuration

Supported File Types

• Images: PNG, JPG, JPEG, GIF, BMP, TIFF
• Documents: PDF (scanned or digital)

Using Variables

Pass file paths from previous nodes:

{{uploaded_file}}
{{scraped_image_url}}
{{document_path}}

Example Use Cases

Process Uploaded Documents

Files Path: {{user_upload.url}}
Output Variable: document_text

Extract Text from Screenshots

Files Path: {{screenshot_url}}
Output Variable: screen_text

Batch Process Images

Use with a For Loop node to process multiple files:

Files Path: {{current_image}}
Output Variable: extracted_text

Output

The extracted text is stored in your output variable:

{{document_text}}

Best Practices

• Use high-quality images for better accuracy
• Ensure good contrast between text and background
• For multi-page PDFs, consider processing pages individually
• Combine with LLM nodes to analyze extracted text