Integration of Z-Depth in compositing

Kayla M Steckel, Purdue University

Abstract

It is important for video compositors to be able to complete their jobs quickly and efficiently. One of the tasks they might encounter is to insert assets such as characters into a 3D rendered environment that has depth information embedded into the image sequence. Currently, a plug-in that facilitates this task (Depth Matte®) functions by looking at the depth information of the layer it's applied to and showing or hiding pixels of that layer. In this plug-in, the Z-Depth used is locked to the layer the plug-in is applied. This research focuses on comparing Depth Matte® to a custom-made plug-in that looks at depth information of a layer other than the one it is applied to, yet showing or hiding the pixels of the layer that it is associated with. Nine subjects tested both Depth Matte® and the custom plug-in ZeDI to gather time and mouse-click data. Time was gathered to test speed and mouse-click data was gathered to test efficiency. ZeDI was shown to be significantly quicker and more efficient, and was also overwhelmingly preferred by the users. In conclusion a technique where pixels are shown dependent on depth information that does not necessarily come from the same layer it's applied to, is quicker and more efficient than one where the depth information is locked to the layer that the plug-in is applied.

Degree

M.S.

Advisors

Whittinghill, Purdue University.

Subject Area

Multimedia Communications|Computer science

COinS