Depth estimation based on global motion and optical flow

摘要

This disclosure describes techniques for estimating a depth of image objects for a two-dimensional (2D) view of a video presentation. For example, an initial indication of depth (e.g., an optical flow) may be determined for a 2D view. The initial indication of depth may be used to estimate global motion, e.g., motion of an observer (e.g., camera), of the 2D view. The initial indication of depth may be modified based on the estimation of global motion to create a global motion-adjusted indication of depth. The global motion-adjusted depth indication may be used to create a depth map for the 2D view, which may be used to generate an alternative view of the video presentation that may be used to display a three-dimensional (3D) video presentation.

主权项

1. A method comprising:
determining, via an image processing module of a computing device for a two-dimensional (2D) view of a video presentation, an optical flow for at least one frame of the 2D view, wherein the optical flow is a representation of an apparent motion of object brightness patterns in a visual plane caused by relative motion between an observer of the 2D view and objects of the 2D view; estimating, by the image processing module and based on the determined optical flow, global motion for the at least one frame; adjusting, by the image processing module, the optical flow based on the estimated global motion to create a global motion-adjusted optical flow for the at least one frame; and generating, by the image processing module, a depth map for the 2D view of the video presentation based on the global motion-adjusted optical flow.