Extract using OCR

How to extract data from documents that require OCR

Extractous supports Optical Character Recognition (OCR) through Tesseract integration, allowing text extraction from scanned documents and images.

Prerequisites

Install Tesseract and required language packs:

# Install Tesseract
sudo apt install tesseract-ocr

# Install language packs (example: German and Arabic)
sudo apt install tesseract-ocr-deu tesseract-ocr-ara

Using OCR with String Extraction

use extractous::Extractor;
use extractous::TesseractOcrConfig;
use extractous::PdfParserConfig;
use extractous::PdfOcrStrategy;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let extractor = Extractor::new()
        .set_ocr_config(
            TesseractOcrConfig::new()
                .set_language("deu")
        )
        .set_pdf_config(
            PdfParserConfig::new()
                .set_ocr_strategy(PdfOcrStrategy::OCR_ONLY)
        );

    let content = extractor.extract_file_to_string("path/to/scanned.pdf")?;
    Ok(())
}

Using OCR with Stream Extraction

use std::io::{BufReader, Read};
use extractous::Extractor;
use extractous::TesseractOcrConfig;
use extractous::PdfParserConfig;
use extractous::PdfOcrStrategy;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let extractor = Extractor::new()
        .set_ocr_config(
            TesseractOcrConfig::new()
                .set_language("deu")
        )
        .set_pdf_config(
            PdfParserConfig::new()
                .set_ocr_strategy(PdfOcrStrategy::OCR_ONLY)
        );

    let stream = extractor.extract_file("path/to/scanned.pdf")?;
    let mut reader = BufReader::new(stream);
    let mut buffer = Vec::new();
    reader.read_to_end(&mut buffer)?;
    Ok(())
}

OCR Strategies

The PdfOcrStrategy enum provides three options:

// Try text extraction first, fall back to OCR if no text found
.set_ocr_strategy(PdfOcrStrategy::OCR_FALLBACK)

// Use OCR exclusively, ignore embedded text
.set_ocr_strategy(PdfOcrStrategy::OCR_ONLY)

// Disable OCR completely
.set_ocr_strategy(PdfOcrStrategy::NO_OCR)

Multi-language Support

Configure OCR for multiple languages:

use extractous::Extractor;
use extractous::TesseractOcrConfig;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    let extractor = Extractor::new()
        .set_ocr_config(
            TesseractOcrConfig::new()
                .set_language("eng+deu+ara")  // Multiple languages
        );

    let content = extractor.extract_file_to_string("path/to/document.pdf")?;
    Ok(())
}
Extract to Stream
How to extract the content from a file (URL/Bytes) to a StreamReader.
Installation
How to install the extractous Python library.