Image Recognition in WinTask
WinTask records/replays using the individual objects seen on the screen, either Windows objects or HTML objects. If an object under automation is not seen, mouse coordinates are used automatically.
Mouse coordinates are not reliable as soon as the automation script has to run on different PCs with different screen resolution. Plus the script is difficult to read and maintain. Sometimes some HTML elements or UI elements can be very stubborn to identify correctly too. For example many UI elements can have the same internal name.