摘要:
Apparatus and methods for picking three dimensional objects from images depicted on a video display. The displayed objects are selectively rerendered. During such rerendering the object pixels are compared in depth to the data in a Z buffer for determining visibility. The number and size of the objects subject to rerendering by the rasterization processor is constrained by using the front end graphics processor to define object extents and by rerendering only extents which have been clipped to the boundaries of the pick window. The rerendering operation does not alter the three dimensional graphics image stored in and repetitively scanned from the frame buffer. Selection between multiple objects within the pick window can include a weighted comparison using a pick plane memory to store visibility data by object.