Hi there, I’m relatively and I’m trying to make kantu do a simple program extract a number from a textbox on discord, an example is shown below:
How would I be able to copy that message and send to verify that captcha?
Any help would be much appreciated! Thank you!
I am pretty sure the new OCRExtract command can read the text and solve this captcha 
OCRExtrat | image_with_green_pink_boxes_relative.png | text
The variable “text” has then the OCR result, which is your captcha. now you can enter it with
xtype | text
Here is how the OCRExtract input image could look like. It finds the image in the green box, and then extracts the text in the pink box:
Oh I see, thank you very much! I was wondering if it isn’t out yet in Mozilla Firefox because I don’t think I see the option for that command yet…
Sorry, I’m still confused on how the OCRExtract actually works… I tried doing the greenbox pinkbox thing but apparently it’s too small and it doesn’t read the text properly? And how would the relative work if multiple instances of the green box appear? If the text is stored in lets say a variable “text,” how would you call it using xtype ( $(text)? or?)? Sorry if these questions are really beginner, I’m really not used to this! Thank you!