searchanddisplace-ingest/app/Ingest/OCR.php

<?php

namespace App\Ingest;

use Illuminate\Support\Facades\File;
use Symfony\Component\Process\Exception\ProcessFailedException;
use Symfony\Component\Process\Process;
use thiagoalessio\TesseractOCR\TesseractOCR;

class OCR
{
    protected $path;

    public function __construct($path)
    {
        $this->path = $path;
    }

    public function execute()
    {
        $this->preProcess();

        $text = $this->extractText();

        return $text;
    }

    protected function preProcess()
    {
        $this->applyDewarp();
        $this->applyDeskew();
    }

    protected function applyDewarp()
    {
        $executablePath = resource_path('python/dewarp/page_dewarp.py');

        $process = new Process([
            'python3',
            $executablePath,
            $this->path,
        ]);

        $process->run();

        if (!$process->isSuccessful()) {
            throw new ProcessFailedException($process);
        }

        $fileName = pathinfo($this->path, PATHINFO_FILENAME);
        $filePath = $fileName . '_thresh.png';
        $directory = pathinfo($this->path, PATHINFO_DIRNAME);
        $newPath = "$directory/$filePath";

        $moved = File::move(base_path($filePath), $newPath);

        if ( ! $moved) {
            throw new \Exception('Something went wrong while moving file.');
        }

        $this->path = $newPath;
    }

    protected function applyDeskew()
    {
        $executablePath = resource_path('libraries/deskew/Bin/deskew');
        $newPath = pathinfo($this->path, PATHINFO_DIRNAME) . '/deskewed.png';

        $process = new Process([
            $executablePath,
            $this->path,
            '-o',
            $newPath
        ]);

        $process->run();

        if ( ! $process->isSuccessful()) {
            throw new ProcessFailedException($process);
        }

        $this->path = $newPath;
    }

    protected function extractText()
    {
        $t = new TesseractOCR($this->path);

//        $t->oem(4);

        $t->psm(4);

        return $t->run();
    }
}