Introduction

What is IRISOCR™ SDK?

The IRISOCR™ SDK (also referred to as iDRS™) is a Software Development Kit (SDK) designed to integrate I.R.I.S. ‘core technologies - such as Optical Character Recognition (OCR), Barcode Recognition (BCR), Intelligent Character Recognition for handprinted characters (ICR), image enhancement, intelligent High-Quality Compression (iHQC), High Quality OCR (HQOCR) and document output technology.

Target audience

The IRISOCR™ SDK is intended for developers who wish to integrate high-quality OCR, BCR, and ICR technologies into their applications and take advantage of the new and improved features offered with the latest version of the SDK.

Compatibility Matrix: Platforms, Languages & Frameworks

The matrix below shows which platforms, languages, and frameworks are supported for integration, helping you quickly identify compatibility for your implementation.

Note

✔ available ✘ NOT available

Table 1. Integration availability by platform/Language or platform/framework

Platform	Windows Intel		Windows ARM	Linux Intel		Linux ARM		macOS	Android	iOS
Bitness	32	64	64	32	64	32	64	universal	32/64	universal
C++	✔	✔	✔	✔	✔	✔	✔	✔	✔	✔
C	✔	✔	✔	✔	✔	✔	✔	✔	✘	✘
.NET	✔	✔	✔	✘	✔	✔	✔	✔	✘	✘

Additional platforms can be supported upon specific project.

The present documentation covers information about integrations for C++ developers or .NET developers.

Main features

Image Input & Output

You can load the image(s) you want to recognize.

More technical information about the supported formats, the color modes and the limitations of each format can be found in the Datasheet.

Caution

The formats list depends on the purchased licenses.

Image enhancement

The image quality is very important for the recognition step.
When the image parameters are not correctly adjusted, e.g. when the image is scanned at a
wrong angle, having low image brightness, being cropped, containing noise or punched
holes, saturation and gamma, then the page recognition is highly affected.

In order to increase the recognition accuracy and also provide a better image compression, Preprocessing corrects the image parameters. Preprocessing comes before the image recognition!

Optical Character Recognition (OCR)

The Optical character recognition not only reads the page but also detects the text, table and graphics zones in order to convert them to editable and searchable text.
The recognition can be performed on page level (full-page text recognition) or on zonal level (zonal text recognition).
It uses advanced recognition algorithms and lexicons in order to recognize 133 languages.

Intelligent Character Recognition (ICR)

The Intelligent Character recognition is used to read a page that contains handprinted characters.

Caution

Limitations The recognition of cursive handwritten characters is not supported.

Note

Modules and their extensions An extension cannot be loaded without its corresponding module. In the case of the ICR extension, the OCR module is required. More information about the modules and the extensions can be found in the Modules section.

Barcode Recognition (BCR)

Danger

Barcode Detection is not available on mobile platforms.

The Barcode recognition is used to perform the recognition of an image containing barcodes.
The recognition can be performed on page level (full-page barcode recognition) or on zonal
level (zonal barcode recognition).

More technical information about the supported barcode types can be found in the Datasheet.

Document Output

After the recognition step, you can save the recognized text as a file in the formats listed in the Datasheet.

Multithread safety

Thanks to this feature, several tasks - or more formally ‘threads’ - can be executed at the same time.

Multithreading reduces the time needed for the execution of different tasks; for instance, barcode/text detection during page processing and PDF conversion.
However, be aware of some constraints: the OCR can recognize many pages at the same time but the Document Output needs the processing results from at least one page before saving them as an output document.

The number of threads to be used depends on your system capabilities.
In the specific case of text recognition, the multithread feature can be used to create several CTextRecognition objects in order to recognize text on many pages.

Furthermore, in certain cases, the multithread feature can be used for faster processing of only one page: one CTextRecognition object may use several threads in order to recognize text in a single page.
Use the methods GetOcrParams() and GetThreadingMode() of the COcrParams class to define the number of threads you want to use.

SDK architecture

The IRISOCR™ SDK is an object-oriented development toolkit. You can find detailed information on properties, methods and classes in the API REFERENCE.