Written by Ana Canteli on 6 july 2017
Lately industry news talks about artificial intelligence systems and recognition systems. The Automatic Digital Recognition is a recognition systems technology based in optical recognition, used in many sector. For example, the Automatic Number Plate recognition - ANPR - is a digital recognition system that uses optical character recognition on images to read vehicles registration plates. The license plate recognition technology is used in real life for access control in parkings, pay-per-use roads, vehicle location or traffic fines.
Digital recognition systems have replaced human repetitive task, reducing identification errors and liberate them - the human - for more value task. OpenKM is a document management systems technology provider, that provide an integrate thirdparty recognition systems focused in real world document management.
Digital recognition systems in document management are focused in type document recognition. Automatic Digital Recognition - ADR - is a recognition technology in document management systems, that makes the application capable of finding information, without being the type of the document a limitation for its identification and management.
In any organization we find different types of documents depending on their structure:
Structured documents: it follows a defined pattern or structure. For example: templates, forms, reports, etc. This means that the information is always in the same place within the document.
Semi-structured documents: follow a similar but not identical structure. For example, delivery notes, invoices, orders, the orders confirmations, etc. In this case, the data can be searched for keywords that are repeated in the documents.
Unstructured documents: each document has its own structure and is different; such as brochures, budgets, technical descriptions, list of requirements, etc. Key data is searched, as in the case of semi-structured documents, thanks to key terms that will appear in the body of the document.
In order to use the Automatic Digital Recognition, within the document management system, documents have to go through the following steps:
Capture: all documents have to be entered in the document management system. Paper documents can be digitized, through the scanner; handwritten documents can be captured by applying a barcode - generated by the document management system. This in turn, can be a solution for large-scale digitization.
Classification: Once the documents are captured and prior to their management, the document is classified, within the document management system, according to the interests of the organization. The documentation can be classified by type, identified by the presence of certain key data: the name of the customer / supplier, date, VAT, etc.
Management: Given the type of document, it can be subject of an automated process that recognizes and extracts the data considered key. For this purpose the DMS uses the Optical Character Recognition (OCR) or Optical Brand Recognition (OMR).
Validation: Sometimes, not all documentation is automatically recognized by the system, due to some failure or error; stains, bends, incorrect rotation angles, etc. In such cases, it is advisable to create a verification environment, in which a user can review the documentation and, once the incident is resolved, forward the document to the management cycle.
Workflows: The documents can be object of workflows that guarantee a univocal application of the procedures, so that the next phase of the workflo is not reached without meeting the requirements of the previous stage.
Delivery: Once the documentation is validated, it must be stored in the target system or application, automatically. For this, it is necessary to have integration facilities, to connect the document management software to third applications.
When the company has the necessary information, but it does not reach the user who needs it; this can lead to loss of opportunities: poor perception of the products and services it provides, delays in payments, meet of commitments, etc.
As an example, a company that manages 500 documents per day (invoices, delivery notes, orders, etc.) and a user who enters the data of these documents in the system at the rate of 30 seconds per document. This implies that the user devotes 15,000 seconds a day; that is to say 250 minutes that suppose more than 4 hours destined to this activity. At the end of the year he will have accumulated a minimum of 960 hours; 20 working days (4 hours x 12 months) during which, with a high probability, they will make mistakes in the insertion process. And the most important; this repetitive work has 2 negative effects: a recurrent cost associated with this process and the loss of intellectual capital within the company.
Automatic Digital Recognition:
It facilitates the transfer of the documentation within the company, minimizing the errors and the risks derived from them.
It allows the identification of the document, which favors the capture of data asmetadata.
It facilitates the search and localization of information over time.
It helps automate the most tedious tasks, freeing staff from highly repetitive tasks; minimizing errors and delays and allowing people to engage in more productive tasks.
It results in costs minimization.
ChronoScan is OCR data capture technology provider. With this formula ChronoScan has managed to capture the most usual expressions in documents: document number, order number, account number, subtotal, VAT, total; without the need to configure templates. Creating these tags it is easier to define the areas to capture or detect metadata.
In addition, the repository import and metadata capture process, can be automated with the OpenKM Import Station, increasing the effectiveness and efficiency of the company's document management.
In the following video you can see an example of using OpenKM integrated with Chronoscan and the Import Station.