Conversion of monoscopic visual content using image-depth database

摘要

An image converter compiles three-dimensional content into a data store, identifies a number of stereo image pairs from the three-dimensional content, computes a depth map for each of the stereo image pairs from the three-dimensional content, and partitions the stereo image pairs in the data store into multiple categories. The image converter determines a depth cue for each of the categories based on the depth map for each of the stereo image pairs in each category. The image converter computes a depth map for a category associated with a two-dimensional input image based on the determined depth cue and renders a three-dimensional output image from the two-dimensional input image using the depth map for the category.

主权项

1. A method comprising:
receiving a two-dimensional image to be converted to a three-dimensional image; determining a category of the image; applying feature values for a plurality of pixels in the two-dimensional image to a feature-to-depth mapping function associated with the category to determine a depth value for each of the plurality of pixels; and generating, by a processing device, the three-dimensional image based on the depth values for the plurality of pixels of the two-dimensional image.