How to convert PDF to Text with Tesseract OCR

pdf_to_text_tesseract_orc.md

How to convert PDF file to text with Tesseract OCR and Imagick

Instalation

Here we need to install the Tesseract and Imagick (to convert PDF to image)

RUN apt-get update && apt-get install -y \
    libmagickwand-dev \
    imagemagick \
    ghostscript \
    tesseract-ocr-por \
    && apt-get autoremove --purge -y && apt-get clean \
    && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
    
RUN pecl install imagick \
    && docker-php-ext-enable imagick
    
RUN sed -i 's/rights="none" pattern="PDF"/rights="read|write" pattern="PDF"/' /etc/ImageMagick-6/policy.xml    

Example using PdfToText class

  $pdf = public_path("cnh.pdf");
  $pdfToText = new PdfToText($pdf);

  $text = $pdfToText->convert();
  
  dump($text);

PdfToText.php

<?php

namespace Infrastructure\Utils\OCR;

use Exception;
use Imagick;
use ImagickException;
use RuntimeException;
use thiagoalessio\TesseractOCR\TesseractOCR;

class PdfToText
{
    private string $pdfPath;
    private string $tempDir;
    private ?int $resolution;

    public function __construct(string $pdfPath, string $tempDir = 'temp', ?int $resolution = 300)
    {
        $this->pdfPath = $pdfPath;
        $this->tempDir = $tempDir;
        $this->resolution = $resolution;

        if (!is_dir($this->tempDir)) {
            mkdir($this->tempDir, 0777, true);
        }
    }

    /**
     * Convert PDF content to image.
     * @param int|null $startPage
     * @param int|null $endPage
     * @return string
     */
    public function convert(int $startPage = null, int $endPage = null): string
    {
        try {
            $imagick = new Imagick();
            $imagick->setResolution($this->resolution, $this->resolution);
            $imagick->readImage($this->pdfPath);

            $allText = [];
            $totalPages = $imagick->getNumberImages();

            $startPage = $startPage ?: 1;
            $endPage = $endPage ?: $totalPages;

            for ($i = $startPage - 1; $i < $endPage; $i++) {
                // set the current page
                $imagick->setIteratorIndex($i);

                // convert to png
                $imagick->setImageFormat('png');

                // improve image quality
                $imagick->transformImageColorspace(Imagick::COLORSPACE_RGB);
                $imagick->normalizeImage();
                $imagick->enhanceImage();
                $imagick->sharpenImage(0, 1);

                // save page to a temp file
                $tempImagePath = sprintf('%s/page_%d.png', $this->tempDir, $i + 1);
                $imagick->writeImage($tempImagePath);

                /*
                 * Process the OCR
                 *
                 * PSM (Page Segmentation Mode):
                 *  0  = Orientation and script detection (OSD) only
                 *  1  = Automatic page segmentation with OSD
                 *  2  = Automatic page segmentation, but no OSD, or OCR
                 *  3  = Fully automatic page segmentation, but no OSD (Default)
                 *  4  = Assume a single column of text of variable sizes
                 *  5  = Assume a single uniform block of vertically aligned text
                 *  6  = Assume a single uniform block of text
                 *  7  = Treat the image as a single text line
                 *  8  = Treat the image as a single word
                 *  9  = Treat the image as a single word in a circle
                 *  10 = Treat the image as a single character
                 *  11 = Sparse text. Find as much text as possible in no particular order
                 *  12 = Sparse text with OSD
                 *  13 = Raw line. Treat the image as a single text line
                 *
                 * OEM (OCR Engine Mode):
                 * 0 = Legacy engine only
                 * 1 = Neural nets LSTM engine only
                 * 2 = Legacy + LSTM engines
                 * 3 = Default, based on what is available
                 */
                $ocr = new TesseractOCR($tempImagePath);
                $ocr->lang('por')
                    ->psm(6)
                    ->oem(1);

                $pageText = $ocr->run();
                $allText[] = sprintf("=== Página %d ===\n%s", $i + 1, $pageText);

                unlink($tempImagePath);
            }

            $imagick->clear();
            $imagick->destroy();

            return implode("\n\n", $allText);

        } catch (ImagickException $e) {
            throw new RuntimeException("Image processing error: " . $e->getMessage());
        } catch (Exception $e) {
            throw new RuntimeException("Error: " . $e->getMessage());
        }
    }

    /**
     * Check if PDF file is protected.
     * @return bool
     */
    public function isProtected(): bool
    {
        try {
            $imagick = new Imagick();
            $imagick->pingImage($this->pdfPath);
            $imagick->destroy();
            return false;
        } catch (ImagickException) {
            return true;
        }
    }

    /**
     * Get PDF info.
     * @return array
     */
    public function getPDFInfo(): array
    {
        try {
            $imagick = new Imagick();
            $imagick->readImage($this->pdfPath);

            $info = [
                'numberOfPages' => $imagick->getNumberImages(),
                'resolution' => $imagick->getImageResolution(),
                'format' => $imagick->getImageFormat(),
                'colorspace' => $imagick->getImageColorspace(),
            ];

            $imagick->destroy();

            return $info;
        } catch (ImagickException $e) {
            throw new RuntimeException("Fail to get PDF info: " . $e->getMessage());
        }
    }
}