Create SearchElement in Abbyy FlexiLayout to find multiple Static text from documents - Manoj Jha

Manoj Jha

Robotic Process Automation, Abbyy Flexicapture, Python, JavaScript, C#. Machine learning and Data Science

Create SearchElement in Abbyy FlexiLayout to find multiple Static text from documents

Share This
            In this post, I am going to explain to you how you can create SearchElement in Abbyy FlexiLayout to detect the various titles of documents, tables, fields, or any text present on all or most of the images in a document.
1.1- FlexiLayout Studio

In Abbyy FlexiLayout, "static text" refers to the text that appears in a specific location on a document and does not change from one document to another. For example, the text "invoice number" on an invoice document would be considered static text because it is always in the same location on the document and its content does not change.

When creating a FlexiLayout template, you can use the software's tools to identify and define the location of static text on a document. This includes things like setting the text's font, size, and location on the page. Once the static text is defined, the software can use this information to automatically recognize and extract the text from similar documents.

The static text definition can be used as the base for recognition of other fields on the document and also can be used to verify the accuracy of the fields recognized using OCR(Optical Character Recognition) or other techniques.

In addition to these, FlexiLayout can also make use of the static text to perform form recognition and validate fields, like date, amount, name etc.

In any document that we want to work on, there are always titles/labels that are the same in all other same types of documents here I mean to say like suppose we are working on some Invoice document to detect the required field example: Invoice Number, Invoice date, Total amount, etc these fields are very common to have in the invoice. These titles have their respective value as a variable (not the same in all invoices) but their titles are always the same means they are static for all the invoices. 
So whenever we start any project in Abbyy our first task is to detect all the static fields existing on the documents. For that purpose, Abbyy FlexiLayout has an element in Studio called Static Text element.

So let's start to understand how we can use Static Text element in FlexiLayout: 
1.2- Static Text property
The above-shown image is the Static Text element. Static Text elements are marked with  in the FlexiLayout tree. The Static text element will consider Recognized words and Recognized lines objects detected during pre-recognition and located in the element's search area as static text candidates. 
To open the Static Text element right-click on the Searchelements after right-click on SearchElements you'll see an option of Add Elements now click on Add Elements now another list will show up in this list you'll see a bunch of elements out of all you've to select one called Static Elements.
When you click on Static Elements a property box will open look exactly the same as above image "1.2- Static Text property". 
Now in this under the General tab first type Element name, it can be anything as per your choice but I would recommend you write the name of the title that you want to detect so it'll be easy to distinguish later. 
After name now click on Static text tab you'll see something like below image:
1.3 - Static Element Property
Now in this tab, you have to write the title/label text in the Search text box, as it is shown. You can see just to detect the description title/header of a table I have added many other relative words in that search box. So now we need to understand why I have added other text to search when we only want a description?
So the answer is, 
  • To generate a hypothesis corresponding to the specified value, For example, if we do not specify the variant 'Description' but only specify the variant 'Desc', the other character won't be included in the hypothesis for the field name. We get invoices from many source companies so, in the invoices, the title/label can have many synonym words of description such as desc or Description: or designation, etc depending on the company to company. You must specify all the possible static text values even they only differ in punctuation marks.
Now click on OK and match your batch image with the search elements. You'll see the items are being detected after all the above steps.
1.4 - Detected Title in Abbyy FlexiLayout




Thanks for reading this all long. I'll be posting more Abbyy and RPA related tutorials and tips soon. If you have any problem related to Abbyy installation or anything then please contact me or comment below so I'll try to resolve your issue. 
Stay tuned to my blog. If you like reading my work please show your support by sharing and following my blog. 

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.

If the post was helpful to you, Please donate us so we can create more useful content

Support by Donate

Pages