In case the web page that we are automating has a Captcha, then we can follow the tips below to handle it.


Option 1


There are many 3rd party applications on the market that solve captchas. These apps can be automated through ProcessRobot/WinAutomation.


Hence, depending on the application, it can be automated either using the Command Line Actions or the UI/Windows Actions.


In this case, we handle the captcha by automating the 3rd party app that does the job.



Option 2


Handling the captcha manually. 

At the time that the captcha appears, the Process can notify the user (Display Message, Send Email, Write Text to File) to take action and manually solve the captcha.


//Then you must add the exception handling to all web actions that could fail because of this Captcha to run this new function and then Continue Process Execution!


Option 3 (May not work for all types of captcha)


In case you are using the Process Robot edition, it has a lot of cognitive features, one of the featured action is "Text Detection" action. This action recognizes text patterns in an Image or PDF file, leveraging the Machine Learning services hosted on the cloud platform. The accuracy mainly depends on the type and quality of Image. Kindly note that one must avail the API key in order to use this feature.


How to detect text in a captcha :


1. Capture or download the Screenshot of the Image using -"Take screenshot on Webpage" action.

2. Send the captured Image to the cloud for detection.


-Additionally please note that during the initial run, the accuracy will be 20-30% over a period of the time the accuracy improves drastically, thanks to Machine Learning :)