Inconsistent in identiying the region between deskstop and laptop using surface automation

[malligac]

Hi We are trying to read values from PDF using surface automation which we successfully automated in desktop machine and when deployed the same process in laptop and tried to automate the same PDF we found that region position is getting changed and unable to read values properly from PDF. Kindly help us to resolve this issue or is there any workaround??

[Denis__Dennehy]

Hello, you may have read the guide on the Portal for interfacing with PDF documents. Firstly, Blue Prism is not an OCR tool - although it does have some limited functionality for such. If you are doing something very limited - where the format of the pdf is always the same - than it might be ok - but for large use of multi-format documents you would need a different ICR solution to feed Blue Prism. Y If you are using Blue Prism for any Surface Automation: Your Blue Prism environments should always be virtually hosted and persistant. They should all be the same resolution - so this would guarantee that a Surface Automation interface you create will work on any robot in the estate. If you cannot guarantee persistance (which you should be able to do...) you need to stop using region editor as the basis for your build, and calculate the location of all the elements you want to use dynamically at runtime using Image Recognition to locate labels etc. There is an exercise in the SA training that shows you how to dynamically use an element based upon runtime location rather than build time location.

[vinoth_kumaryuv]

Adding one more point or observation to above query , I have captured the region in PDF using surface automation in one screen resolution (1366*768) and able to read values from PDF Then I changed the screen resolution (1024*768) to check a scenario and after coming back to original resolution (1366*768) we found that region position is getting changed and unable to read values properly from PDF Please let us know what is the cause for this issue and how to resolve the same. Do I need to recapture the region again from PDF?

[vinoth_kumaryuv]

Hi Denis Adding one more observation, I also used dynamic region to locate the elements at runtime using Image Recognition. It was able to read values from PDF when the PDF is in same size but when the PDF size is maximized (with Zoom also automatically getting increased) then image recognition is unable to identify the sub image and consequently again the values are not properly read from the PDF. Kindly help us in resolving this issue with any other work around.