Optical Character Recognition (OCR)

Overview

FormKiQ's Optical Character Recognition (OCR) capabilities transform images and scanned documents into searchable, editable text. This feature enhances document accessibility and enables powerful search functionality across your document repository.

OCR Providers

FormKiQ supports two powerful OCR solutions:

Amazon Textract

• Advanced ML-powered OCR service
• Handles complex document formats
• Available in FormKiQ Essentials, Advanced, and Enterprise editions
• Supports automated form field extraction
• Handles tables and structured data
• Processes handwritten text

Tesseract OCR

• Open-source OCR engine by Google
• Available in all FormKiQ editions, including Core
• Ideal for basic text extraction
• Supports multiple languages
• Lightweight and efficient

Supported Document Types

Feature	Tesseract	Amazon Textract
PDF	✓	✓
JPEG	✓	✓
TIFF	✓	✓
PNG	✓	✓
GIF	✓
WEBP	✓
BMP	✓
DOC	✓
DOCX	✓
XLS	✓
XLSX	✓

OCR Configuration

Document Action Parameters

{
  "ocrEngine": "tesseract|textract",
  "ocrParseTypes": ["TEXT", "FORMS", "TABLES"],
  "ocrExportToCsv": true,
  "ocrNumberOfPages": -1,
  "addPdfDetectedCharactersAsText": true
}

Parameter Details

Parameter	Description	Default	Notes
ocrEngine	OCR provider to use	tesseract	Choose between 'tesseract' or 'textract'
ocrParseTypes	Types of content to extract	["TEXT"]	Options: TEXT, FORMS, TABLES
ocrExportToCsv	Export table data to CSV	false	Only applicable for textract table extraction
ocrNumberOfPages	Number of pages to process	-1	-1 processes all pages
addPdfDetectedCharactersAsText	Convert PDF image text	false	Enhances PDF processing

Features by Provider

Tesseract Features

• Basic text extraction
• Multi-language support
• Page segmentation
• Character recognition
• PDF text layer generation

Amazon Textract Features

• Advanced text extraction
• Form field detection
• Table structure recognition
• Key-value pair identification
• Handwriting recognition
• Document layout analysis

Best Practices

1. Document Preparation

   • Ensure good image quality
   • Use appropriate resolution (300 DPI recommended)
   • Properly align documents
   • Remove unnecessary backgrounds

2. Provider Selection

   • Use Textract for complex documents
   • Choose Tesseract for basic text extraction
   • Consider volume requirements
   • Evaluate accuracy needs

3. Performance Optimization

   • Process appropriate page ranges
   • Use specific parse types
   • Enable CSV export only when needed
   • Configure appropriate timeout values

4. Error Handling

   • Implement retry mechanisms
   • Monitor OCR quality
   • Validate output format
   • Handle failed conversions

Common Use Cases

1. Document Digitization

   • Convert paper documents to searchable PDFs
   • Extract text from scanned images
   • Create searchable document archives

2. Form Processing

   • Extract form fields automatically
   • Process structured documents
   • Validate form data

3. Data Extraction

   • Pull information from tables
   • Extract key-value pairs
   • Process invoice data

4. Content Search

   • Enable full-text search
   • Index document content
   • Improve document discovery

API Integration

// Example: Configure OCR Document Action
POST /documents/{documentId}/actions
{
  "action": "OCR",
  "parameters": {
    "ocrEngine": "textract",
    "ocrParseTypes": ["TEXT", "TABLES"],
    "ocrExportToCsv": true,
    "ocrNumberOfPages": 10
  }
}

For complete API documentation, see Document Actions API Reference.