ABBYY FlexiCapture Vs FlexiLayout Discrepancies


I am using FlexiLayout 12 and FlexiCapture 12

I created a FlexiLayout project, and it works great in the FlexiLayout Editor. All elements and blocks are working. When I export the AFL, I go into FlexiLayout to create a document definition. At first, the document definition fails to collect all the data when I run the test for some reason, and it mismatches vs the FlexiLayout project. However, when I check the "Use Recognition Settings from Document Definition" under the document section properties, the document definition collects the right data and matches the FlexiLayout.

Then I publish it. Now, when I use my image import profile or a default batch, FlexiCapture always fails to export and recognize the correct data. I've tried toggling all the different recognition options (fast, balanced, full) and I've messed around with the image enhancement section. At this point I'm just guessing and trying new options. I've also tried the tips from this article:  . However, Black/white was already unchecked in both FlexiLayout and FlexiCapture, and I mismatch. When I try checking Black/White conversion, I still mismatch; in fact it's way worse.

Can someone please offer me some tips? This is frustrating because I've had this happen on many occasions of mine. From reading other posts, it seems that FC and FL have different preprocessing and recognition settings. I thought I checked that these were the same multiple times, but maybe I did it wrong. How can I adjust these settings so my FlexiLayout matches its export? All I want to do is take data out of a PDF and put it into xls just like how I can see it it in my FlexiLayout or is there a way I can just export what I see in the document definition editor??





  • Avatar



    Can you check in flexilayout that you matched correctly the blocks with their corresponding fields?

    Might be also that you changed the data type in the definition editor for the fields?


    That's the only two options I see that could change your output



  • Avatar

    Hi JF,

    Thanks for the response. Actually I think I solved this issue. The blocks and data types were all fine. 

    To solve it, I tried two different things. Unfortunately, I'm not sure which change fixed it or if both changes were required but,

       1)  In the batch type and my image import profile, I tried using the OCR option instead of adding the PDF text layer option. I would say that doing this fixed some other cases I've had where this issue came up.

       2) I also tried just creating a new project. I might have had some weird settings in my original project. So right now I have 1 project where I wish to keep all my document definitions, and If that fails for whatever reason, I will try it in the new project I've made.

    I'm leaning more towards 1) being the solution because like I said, it solved some other cases where this problem came up.



Please sign in to leave a comment.