ahmedghanem00/tesseract-ocr

A PHP wrapper for Tesseract-OCR binary

1.0.0 2023-06-01 22:18 UTC

Requires

• php: >=8.1
• ext-gd: *
• intervention/image: 2.7.*
• symfony/process: 6.*
• symfony/string: 6.*

Requires (Dev)

• ext-imagick: *
• phpstan/phpstan: 1.10.x-dev
• phpunit/phpunit: dev-main
• smalot/pdfparser: dev-master

MIT

• Ahmed Ghanem

• Tesseract
• OCR
• text recognition
• php-wrapper
• ocr recognition

README

A PHP wrapper for Tesseract-OCR binary.

Originally inspired from ddeboer/tesseract with added features + some Improvements.

Installation

$ composer require ahmedghanem00/tesseract-ocr

Usage

if the tesseract is added to your path, You can just do:

$tesseract = new \ahmedghanem00\TesseractOCR\Tesseract();

Otherwise, You can do:

$tesseract = new \ahmedghanem00\TesseractOCR\Tesseract("path/to/binary/location");
# OR
$tesseract->setBinaryPath("path/to/binary/location");

To specify the tesseract process timeout:

$tesseract = new \ahmedghanem00\TesseractOCR\Tesseract(processTimeout: 3);
# OR
$tesseract->setProcessTimeout(2.5);

To specify a custom tessdata-dir:

$tesseract->setTessDataDirPath("path/to/data/dir")

To reset tessdata-dir to default:

$tesseract->resetTessDataDirPath();

To get version of the binary:

$version = $tesseract->getVersion();

To get all the supported languages:

$languages = $tesseract->getSupportedLanguages();

To OCR an Image:

$result = $tesseract->recognize("test.png");
##
$result = $tesseract->recognize("https://example.com/test.png");

Thanks to the Intervention/image package. The recognize method can accept different sources for an image:

- Path of the image in filesystem.
- URL of an image (allow_url_fopen must be enabled).
- Binary image data.
- Data-URL encoded image data.
- Base64 encoded image data.
- PHP resource of type gd
- Imagick instance
- Intervention\Image\Image instance
- SplFileInfo instance (To handle Laravel file uploads via Symfony\Component\HttpFoundation\File\UploadedFile)

To Specify the language(s):

$result = $tesseract->recognize("test.png", langs: ["eng", "ara"]);

To specify the Page-Segmentation-Model (PSM):

use ahmedghanem00\TesseractOCR\Enum\PSM;

# using PSM enum
$result = $tesseract->recognize("test.png", psm: PSM::SINGLE_BLOCK);
# OR by using id directly
$result = $tesseract->recognize("test.png", psm: 3);

To specify the OCR-Engine-Mode (OEM):

use ahmedghanem00\TesseractOCR\Enum\OEM;

# using OEM enum
$result = $tesseract->recognize("test.png", oem: OEM::LEGACY_WITH_LSTM);
# OR by using id directly
$result = $tesseract->recognize("test.png", oem: 3);

To specify the DPI of the input image:

$result = $tesseract->recognize("test.png", dpi: 200);

To make the recognize method output the result as a searchable PDF instead of raw text:

$pdfBinaryData = $tesseract->recognize("test.png", outputAsPDF: true);

file_put_contents("result.pdf", $pdfBinaryData)

To specify words-file or patterns-file:

$result = $tesseract->recognize("test.png", wordsFilePath: "/path/to/file");
# OR
$result = $tesseract->recognize("test.png", patternsFilePath: "/path/to/file");

To set a config parameters:

use ahmedghanem00\TesseractOCR\ConfigBag;

$config = ConfigBag::new()
    ->setParameter("tessedit_char_whitelist", "abcrety")
    ->setParameter("textord_pitch_range", 3);

$result = $tesseract->recognize("test.png", config: $config);

You can also run tesseract --print-parameters to see the list of available config parameters.

Licence

tesseract-ocr is licensed under the MIT License.