README

This package is a wrapper around the the tesseract-ocr (https://github.com/tesseract-ocr/tesseract) package. This library depends on

Symfony Process
Tesseract OCR, version 4 or later.

Installation

composer require dbizapps/ocr

Dependencies

There are many options to install Tesseract OCR according to your opeerating system needs. Personal experiences are limited to macOS for which you can find a recommendation below. Sorry for this inconveniences.

macOS

To install Tesseract OCR with support just for english

$ brew install tesseract

For a fully fledged language installation you can choose

$ brew install tesseract tesseract-lang

Basic Usage

Creating a recognition engine instance doesn't require further attributes

// create ocr instance
$engine = new Recognition;

Retrieving tesseract specific information

// retrieve tesseract version
$version = $engine->version();

// retrieve supported languages
$lang = $engine->supportedLanguages();

Input/Output Configuration

The target image can be set while creating a new recognition instance or at a later state

$engine = new Recognition(<image-path>);

$engine->setImage(<image-path>);

The engine can act in file mode to store results in a given output file and output path Without providing output configuration options the later scan process is streaming results

$engine = new Recognition(<image-path>, <output-dir>, <output-filename>);

Custom Configuration

To get an overview of Tesseract OCR configuration options, please refer to https://github.com/tesseract-ocr/tesseract

Configuration options can be set individually like

// Scan density
$engine->setConfig(['dpi' => 300]);

// OCR Engine mode.
$engine->setConfig(['oem' => 2]);

// Page segmentation method
$engine->setConfig(['psm' => 10]);

// Languages 
$engine->setConfig(['languages' => ['eng', 'fra']]);

// output configuration
$engine->setConfig(['configFile' => ['pdf']]);

Supported options for configFile are:

alto Output in ALTO format (OUTPUTBASE.xml).
hocr Output in hOCR format (OUTPUTBASE.hocr).
pdf Output PDF (OUTPUTBASE.pdf).
txt Output plain text (OUTPUTBASE.txt / default).

Instead of setting options individually there is a global settings method

$engine->setConfig([
	'dpi' => 300,
	'oem' => 2,
	'psm' => 10,
	'languages' => ['eng', 'fra'],
	'configFile' => ['pdf'],
])

Scan usage

Scanning an image as simple as

$result = $engine->scan()

If the output-path and output-filename are set, the provided result represents the path to the scaned output. Otherwise the result respresents the scaned output.

timatanga / ocr

Maintainers

Details