Advanced Optical Character Recognition

Advanced Optical Character Recognition

Written by Ana Canteli on 13 May 2022

Last week we discussed the zonal OCR engine included as standard in OpenKM. It intervenes in many tasks that make document management very easy, fast, and affordable for the user in his daily work. From extracting the text of documents, a person can find any document based on any word they remember; to enable the identification of document types, indexing, categorization, and assignment of metadata groups automatically, to name just a few examples.

OpenKM's document management software can be integrated with any third-party application thanks to the web services and the complete API in its documentation. Today we will talk about the possibility of using OpenKM with an advanced optical character recognition application.

Advanced zonal OCR expands the number of cases in which optical character recognition can be successfully implemented in the document processes and business processes that the organization wishes.

Benefits of Optical Character Recognition

OCR provides tangible benefits, primarily when used for specific tasks:

• Facilitates accounting: optical character recognition is often used to speed up the processing of documentation that reaches the company in large quantities and affects many processes. Invoice management, for example, is not only involved in the accounting department but is also relevant at the level of customer service, financial department, after-sales service management, warehouse, etc.

• Reduction of waiting times: the manual or labor-intensive processing of regular documentation implies time-lapses that are impossible to reduce. Thanks to the use of OCR, essential documentation management can be automated so that, for example, the payment of customer and supplier invoices is streamlined.

• Minimization of errors: subjecting human resources to performing repetitive tasks with little added value increases the possibility of incurring human mistakes over time, which can be avoided for the most part with the use of optical character recognition.

• Maximization of utility: Faster management cycles and reduction of errors, together with the minimization of the impact that these can have on the daily administration of the business, implies the optimization of document management throughout the company.

• Greater customer satisfaction: and of all the parties involved with the organization, such as suppliers or public administrations. When the company enables access to its products and services through extended document management, third parties can benefit from modular access to company information and documentation. If the platform is optimized thanks to optical character recognition, the experience offered will be more satisfactory.

Advanced Optical Character Recognition

OpenKM offers an advanced optical character recognition solution. This option enables the management of large volumes of documentation. Although they are the same type of document (for example, invoices), they are not limited by the design differences of each one. of them. It must be taken into account that each invoice that an organization receives comes from a different entity, and therefore the critical values ​​of this type of file can be found in various places within the same document. For open-source OCR engines, this represents a limitation that can only be overcome by inserting templates, in which the OCR application is told where the values ​​it has to read are.

Advanced optical character recognition allows the capture of metadata such as date, invoice number, concept, units, price, tax base, VAT, subtotal, total, etc., without the design of each invoice being a limitation. Such an advantage multiplies the possibilities of automated document management for the company.

Advanced optical character recognition use case

Here we present a standard use case of invoice management using OpenKM integrated with an advanced OCR solution.

In the organization, there will be an app where the documentation will be generated or recorded; delivery notes, invoices, tickets, purchase orders, orders, etc.

We start with invoices, particularly supplier invoices. The documentation is deposited in OpenKM.

The documentation to be analyzed can come from different sources and other formats. As an email attachment or as a result of a scan, it will be a single file made up of several pages, each of which would be an invoice.

In OpenKM, a documentation entry route and an exit route are programmed. In the simplest scenario, an input file - a pdf, for example - will contain one single sheet, an invoice. Usually, an input document can have several exit points, so we can take a folder per client when preparing the taxonomy. There will be a subfolder as the entry point of the documentation. The invoices are organized by date (year/month/day) in this folder.

This advanced optical character recognition system includes configuration services within OpenKM. The technical team will be in charge of implementing the document input route, the import service, error control (what is the action protocol in this case), and document separation, among other features.

When the input folder hosts documentation, OpenKM will transfer the documentation to advanced OCR - capture, processing, separation - and the documentation goes to the output path, where the documentation will be ready for use.

If the advanced optical character recognition application detects any errors, the user will receive a notification with the wrong or defective file. It can also be hosted in a folder in OpenKM, where the user can manage the incident. All these operations are usually carried out in the background not to affect the system's performance for the user.

This represents just one case of many scenarios in which advanced optical character recognition can improve your company's day-to-day document management. Contact us for more information.