Currenlty we have series of Remote Desktop (Citrix) based automation to be performed using UiPath. As we got less accurate results using Tesseract and Microsoft OCR, Now we have subscribed to Abbyy OCR. Somehow using OCR we are getting results but when screen position changes the values are not accurate every time. Here due to narrow timelines we are now close to Deadline to complete these Citrix based automations.
So request here to kindly connect with us and provide feasible solution whether to use ABBYY / CV based on our Business Scenario and let us know the strategy to take forward this situation. Few live examples of things happening at our machine using Abbyy/CV has been created.
We are unable to use Abbyy fully to scrap screen data, we are using default OCR engine for relative scrapping and then changing abbyy to get the value against label. The default OCR engine to identify label is giving inconsistent results and unable to identify correct label, hence abbyy returning a different value.
for e.g. when trying to read “Primary Servicing Producer ID:”, the OCR engine sometimes returning value for “Primary Servicing Producer” as highlighted in red (#1 & #2). If we trying to find only “Producer ID:”, it is sometimes returning the value for “Producer Identifier” as highlighted in green (#3).
We would like to try Computer Vision to identify correct label, kindly suggest if we can get on-premise version of CV. Also, suggest us if we can raise this issue with UiPath or Abbyy?