What is OCR
OCR (optical character recognition) is the translation of various types of images (handwritten, typewritten, printed and others) into a textual electronic representation for further editing, processing and analysis of information. In a simple concept, this is the process of obtaining text from an image.
In software representation, OCR is an image processing algorithm developed in a particular programming language that programmatically recognizes text characters.
Such software products are used to digitize books, newspapers, magazines, various text documentation, archival information, and so on.
OCR and WEB
As part of WEB OCR systems can be used for a variety of purposes. As an example, such a system can be connected to the site and used to process waybills (in the scanned version) in order to automate the process of accounting for goods on the site. Also, this functionality can be used in online stores to provide the client with a link to track the location of the parcel, if the carrier transport company offers such functionality. Such solutions can be implemented by recognizing the necessary data from the waybill of the carrier company and creating a link for the client, by clicking on which he can see the status and position of the cargo sent for him.
Existing software
Currently, there are quite a few options for the software implementation of OCR systems, which differ in many parameters and features: from algorithms, the subject of recognized text, the availability of a desktop program, online availability, and ending with the price per volume of a recognizable unit. But, if we consider the use of OCR within the WEB, we can distinguish several services that are distinguished by ease of use, quality of services provided and affordability.
ABBYY Cloud OCR SDK
The first of the considered OCR systems is ABBYY Cloud OCR SDK. ABBY itself, the author of this development, is famous for its computer program ABYY FineReader, which is widely used in various fields, from scientific and educational to use in private companies, and is still being developed and maintained.
The recent development of ABBYY Cloud OCR SDK, a text recognition service, makes it possible to embed this system into your website. On official website a brief description of the system itself, tariffs and terms of use, examples of software implementation, a demo version of their service, and so on.
The implementation itself on the site is not particularly difficult, since ready-made examples can be safely inserted into the code and, having filled in the necessary security and identification data, use the service.
ABBYY Cloud OCR SDK is distinguished by a wide choice of languages for representing the resulting text (about 198) and the quality of translation.
If we talk about the terms of use, then initially 50 sheets of A4 format recognizable text are provided free of charge. After using them, various packages are offered, allowing you to buy an increase in the limit to 100 sheets for 9.99$, to 1000 sheets for 69.99 $, to 5 thousand - 199.99 $, to 30 thousand - 899.99 $ and, if necessary, more 100 thousand sheets, then you need to contact the sales department.
After image recognition, ABBYY Cloud OCR SDK offers the ability to output text in TXT, DOCX, XLSX, PDF, XML, CSV and other formats. This choice of formats helps to facilitate the processing of the resulting information, receiving it in a convenient or necessary format.
OCR WEB SERVICE
The following service provides similar functionality to ABBYY Cloud OCR SDK. The difference is the software implementation, examples of which are provided on the .NET software platform and programming languages JAVA and PHP.
The number of available languages is limited to 46.
The quality of the Latin alphabet text recognition is at a high level, but there are often problems with the Cyrillic alphabet (when the service returns incomprehensible characters as a result of work). You can read about these issues on various official and unofficial forums.
If we talk about price characteristics, then, for starters, the service gives a month of free use with a limit of 25 pages per day. Further, the tariff plans for use differ from the time of use or the number of resulting sheets. So, 1000 pages of recognized text cost 39.95 $.
The recognition result can be output in popular formats like TXT, DOC, XLS, PDF and others.
Official site provides all the detailed information and terms of use.
Free OCR API
This service provides simple and inexpensive (compared to previous options) functionality. An example and implementation description is provided in PHP. Approximately 70 languages of recognizable text are supported.
The result of the work is displayed as a stream of string data, which, for further data processing, requires understanding and experience in using the PHP programming language.
After activating the account for using the service, it is possible to recognize 200 images (pages) for free. At the end of this amount, each subsequent image costs 0.5 $.
Summarizing
Despite the complexity of the concept and implementation of OCR, such systems and services are currently widely used and used for commercial and other purposes. The development and growth of the popularity of the considered services only confirms this fact and provides an opportunity to relatively easily use this technology when developing your site.
Author: web-programmer of the agency of strategic Internet marketing Star Marketing Yaroslav Kustrich.
