ABBYY FlexiCapture FAQs
Whereas our featured OCR software reviews provide real-life feedback and case studies from actual ABBYY FlexiCapture customers, this page seeks to address some of the most common questions about when, why, and how to leverage OCR software for data capture.
The goal of OCR software is simple—to transform typed text into usable data—but the algorithms are highly complex. To simplify for the sake of explanation, OCR software usually follows a text recognition process like the following:
1) The software detects the boundaries of a single character (i.e., draws a sort of imaginary box)
2) It calculates which character it most likely sees, based on the location of strokes/marks within that area and the characters next to it
3) It stores that character with a certainty level, then proceeds to the next one, and so forth until the end of the document
4) After recognizing all characters, a human operator confirms the ones with certainty levels below the customer's preferred threshold
5) Once confirmed, the newly digitized information is sent to a database, CSV file, or virtually any other destination for storage or further processing and/or validation
For accuracy, the precision of the recognition algorithm is everything. ABBYY FlexiCapture achieves industry-leading accuracy thanks to continuing R&D by a large team of computer scientists, linguists, and—literally—rocket scientists. Accuracy rates of around 99.9% are common when the software has been properly implemented. In our experience, these are the best OCR software results currently possible.
FlexiCapture does have a built-in learning ability, which means results can improve as it ‘sees’ more and more of a certain document. Similarly, any documents that are recognized on a regular basis should be set up as a template, which makes the software even more accurate by telling it which characters to expect.
Yes, although in that case it’s not called optical character recognition; instead, it's referred to as intelligent character recognition (ICR).
Handwriting is notoriously difficult to process given its inherent inconsistency, variation in spacing between letters, and other challenges. However, properly designed forms and a high-precision OCR application such as FlexiCapture can still return accurate and useful results. This is particularly valuable for processes like patient registration automation and customs / logistics document processing, to name a couple.
The exact improvements depend on the length, complexity, and variety of documents. That said, a 75% reduction in processing time and cost is a common but conservative figure. Some organizations have been able to reduce data entry FTEs by upwards of 90% after deploying optical character recognition software, with commensurate reduction in costs.
Even though efficiency is usually thought of in terms of time and cost, accuracy is another critical dimension. Baseline error rates depend completely on the training and ability of the data entry specialists, but as mentioned earlier, FlexiCapture is capable of 99.9% accuracy. For almost any organization, this represents a substantial improvement over manual data entry.
Not remotely. First off, our team does the hard part for you. Configuration, template/form creation, support, training, and everything else are routinely provided as part of your comprehensive data capture automation solution.
We design each solution in close cooperation with the customer, to ensure that it adds value and removes obstacles at every step of the workflow. The result is a straightforward and reliable data capture process, with practically no human involvement beyond a single keystroke to confirm/correct ambiguous characters.
Right here! To buy FlexiCapture and/or see pricing for the full range of FlexiCapture licenses, simply visit our OCR Data Capture page and create an account. As always, please contact us with any questions or comments. Our team is here to help you understand your options and choose the best OCR software for your technical and financial needs.