AutoIt: any support for OCR?

[anukul]

Hi,

I am automating my set top box, having a Client app which has a remote which synchronizes with the TV at the time it connects. (after having entered STB IP and box type). After successfully connecting, I can simulate all the actions on the TV with the remote present in my client app (they go in parallel). For e.g, if I press "menu" on the remote in my client app, "menu" click is simulated on the TV as well. Also, the screen shown on TV is also shown in the client app.

Coming to the question, my requirement is to capture the image coming on my client app (basically an image capture). I have a tool QTP which supports Optical Character Recognition (OCR) and hence I am able to create checkpoints and get text out of the image.

I am now using ruby with AutoItX controls to get this client app automated.It seems to be working really well and fast for me, just facing one problem to fetch text out of the screen image.

But my question is, does AutoIt support getting text out of images? Is there any way out to work this?

Thanks,

Anukul

[mafioso]

It is possible, but not in pure AutoIt3. This has been discussed numerous times and you should search the forums. Using Microsoft Office's built-in OCR, it is possible to get text out of the images. But the result may not be always good.

[Yorn]

I've found the microsoft OCR is actually *very* good. The problem is you have to have Office 2000, XP, 2003 or 2007 installed, and you have to pay for those.

[Andreik]

I've found the microsoft OCR is actually *very* good. The problem is you have to have Office 2000, XP, 2003 or 2007 installed, and you have to pay for those.

Yes, look at ptrex example.