Arrange a consultation between the filmmaker and the technical team. Typically, a filmmaker consults with visual artists and technicians about how images should be composed, what the key objects are in each frame and what the desired perspective for the viewer is.
Begin key-framing, or the process of identifying key objects that will be the primary depth regions of the image (for example, foreground and background). These are often the central characters, key objects or reoccurring scenery. The more details that are key-framed during this process the more complex and deep the images on screen will be. The sacrifice, however, is reducing the resolution on each distinct layer.
Isolate these key objects using key-framing software to create a 3-D depth map overlaying the digitalized 2-D image.
Mask the footage, or extend the depth map, to the entirety of the material. During this process, use image recognition software (looking for shape and luminosity) to apply the rules determined during the key-framing process. This requires a great deal of computing power and significant server space. The final product is a depth map of the entire 3-D film.
Create the 3-D images using 3-D imaging software. Use the depth map conceptualized during the masking stage as a blueprint, but know that there is a great deal of creative freedom during this process. Specific objects are given an assigned volume and place within the visual plane. Effectively this pulls some parts of a flat image forward and pushes other ones back, creating a sense of foreground and background. The filmmaker and 3-D development team should consult regularly during this process.
Remove the remaining artifacts. These are the spaces or gaps left behind when a given image is placed on another plane (for example, closer to or further away from the viewer). Using CGI, fill in the missing spaces. Usually this is a compilation of imagery surrounding the gap to create a reconstruction of the natural image.