Since screen scraping is a delicate technology, not a single approach is enough to cope with various technologies used to create the user interface.

WordCaptureX employs 4 methods to extract text from the screen.

Native screen scraping method

Works by redirecting GDI text rendering functions. Sounds easy, isn't it?

We've spent 4 years of continuous development to make it work flawlessly. Accuracy 100%, works with GDI apps (x32 and x64).

Screen OCR method

It uses Google Tesseract OCR engine under the hood. Major work involved to tweak OCR engine to work with screen fonts. It is supposed to be used where Native method is not working. Accuracy is about 95% for regular fonts.

FullText method

Aims to work with the document object model exposed by the technology used to create the app that you want to scrape. Limited implementation in WordCaptureX.

For complete support see UIElement library of Uipath Studio.

Used best with Flash and FoxPro apps. Accuracy 100% but keep in mind that the full text of the control pointed by the user is retrieved, not just the pointed word. Left context, right context and paragraph properties are not set.

SelectedText method

Used to extract the selected text on any apps. Works best when setting a hotkey to trigger this method. For example a dictionary can choose to allow people to CTRL-RClick on a word to get its definition and also get the definition of a selected text when they press a defined hotkey like CTRL-SHIFT-S.