I have a PDF file which will contains some data like below structure.

I want to use Azure Form Recognizer to get the data.

How can I set the label with Table.

While tagging with Table, it need to specify the Column and Row.

CodePudding user response：

You must select a table from the Form recognizer tag insertion field. The important thing here is to choose the table type (fixed sized or row dynamic).

Then you must point to the fields in the table by manually creating the columns. If there are no columns, I recommend you to label the fields one by one, or you can create imaginary columns and then delete these imaginary columns by writing a script.

You can find more detailed information at the link below.

Use table tags to train your custom template model

CodePudding user response：

Form Recognizer should extract this table automatically as part of the layout information extract from a document. Did Layout succeed to extract the table ? Do you also want to structure it in a certain structure ? If yes, you can also train a custom model (template) and label the table. Labeling the table can be done in any format. That table you label does not need to be in the same structure. For example You can create a 4 column table form this data with Age, Name, Weight, Hight as the columns or if this is a single row table you can also label it as key value pairs. Try out the Form Recognizer Generic Document pre-built it might extract these as key value pairs out of the box - https://formrecognizer.appliedai.azure.com/studio/document