Working together for accurate results
All applications that convert text from an image to an editable or searchable format rely on good quality input. The StatementReader team acknowledge this and work with each client to ensure that an accurate solution is found. This is often a combined effort, here's how.
You
In cases where you have the original paper document, you have control over ensuring that the scanned image is good quality.
To scan documents for the most accurate character recognition, you should:
-
align the paper correctly in the scanner
-
create a searchable PDF document (if possible) or black and white TIF image
-
set the scanner to 300DPI
-
if scanning to a TIF image, select text under image type, so the scanner knows to enhance the contrast and remove any background shading
-
ensure pages are clean and free of pen markings
Us
StatementReader has been specifically developed for financial documents. This means that we:
-
create a monthly summary, for the user to quickly sense check the data
-
use data types to automatically correct dates and numbers
-
use tested templates
-
make page adjustments based on text location
And if you still do not get the output that you require, you can either upload your scanned files to our secure server, or we can arrange a courier to collect your paper documents and we will process the output for you, at no extra cost. Click here for more details.
Together
-
Instant and accurate data conversion from scanned bank statements to Excel
-
Unlimited pages, processed immediately and securely
-
Full application and job support, when you need it
Terminology
Searchable PDF: Office scanners often have the option to create a searchable PDF document. This means that the scanner uses the original image to create a 'text layer' within the PDF document, which StatementReader can extract. We have seen that using the original document to recognise the text often produces far better results than relying on a scan.