Preface
In general, document capture refers to the process of scanning paper documents using scanners or cameras, and transforming these documents into an electronic file such as a PDF, TIFF, and so on. Through a type of document capture software, these electronic files are assessed through character or pattern recognition (that is, OCR, ICR, and OMR) and converted into meaningful data or information, also called metadata. The goal of document capture systems is to classify incoming documents into categories and extract metadata by automating processes that humans normally do. By automating this process, organizations can classify documents and route the incoming documents to repositories and workflow systems with all the metadata more efficiently; thus, document capture systems help reduce errors, allow documents to be handled faster, and organizations can scale their business using software rather than labor.
Over the years, the document capture industry has evolved in many ways. Paper documents are still a big portion of the document formats that organizations receive. However, as organizations started to exchange documents electronically, the document capture systems had to evolve to be able to process the documents within e-mail attachments, sent via FTP or by using APIs. Of course, smartphone adoption also influenced the document capture world. As consumers started to produce electronic documents more using their phones, such as check deposits or expense reports, capture using mobile devices became essential.
The capabilities of the capture software has also become more intelligent over the years. Early systems automated the capture process by using barcodes for classifying documents and extracting metadata from predefined areas of structured documents, called forms. Later on, technologies such as Ephesoft classified the documents using several techniques such as document layout or words and phrases that can extract the metadata without defining where the metadata might be located on the document. We call these capture systems Intelligent Document Capture systems and they can help organizations automate document capture systems not only for structured documents but also for unstructured documents such as e-mails or documents that do not have any known format. Intelligent Document Capture systems are also easier to deploy and maintain compared to older systems, which provides faster ROI and adoption.
The latest trend in document capture systems is the use of web-based APIs. In the beginning, document capture systems were used only to ingest documents to repositories and workflow systems but with the popularity of APIs and the cloud, a new use case has opened. The newest versions of document capture systems now allow organizations to enable other business applications with document capture functionalities, where the cloud, private or public, based APIs provide document capture services rather than the applications. This makes organizations more efficient when it comes to maintaining, upgrading, and integrating multiple business applications.
If we try to imagine where document capture will be 5 to 10 years from now, it is anyone's guess. However, looking at the emerging technologies and today's advancements, we may be able to predict the future. In the future, we will see machine learning algorithms doing what capture administrators do today, so that any user can simply show the next generation document capture system what he/she wants within that document. Then, the computer should be able to understand what to do the next time without any human intervention. What this means is, if it takes one administrator to configure the system today and one IT professional to set up the servers, in the future there will be no need for either resource. The users will simply request a capture service using a browser or mobile device, the service will learn the human behavior by simply observing the user and then will perform the same task from then on. The projects on which we used to spend months on implementing are now measured in weeks. In the future, they will be measured in hours, if not minutes and seconds.
