Systems and methods for rendering 3D image independent of display size and viewing distance

摘要

Methods and systems for providing and rendering 3D depth information are described. Specifically, the 3D depth information includes z-axis values provided in a normalized percentage format, that defines the position of an object relative to an infinity plane, a display screen, and a viewer, and can be used to render one or more images independent of display screen size and viewing distance.

主权项

1. A method of rendering 3D images, comprising:
receiving a profile comprising a profile header, from a source device and independent of display screen size, a depth map, and 3D depth metadata, wherein the 3D depth metadata comprises data for multiple target display resolutions from a source device and said 3D depth metadata is independent of display screen size; using the depth map to derive depth data that is independent of display screen size and viewing distance, wherein the depth data comprises normalized z-axis values based on a plurality of depth points, the normalized z-axis values defining the relative position of an imaged object with respect to a human viewer, wherein the profile header comprises a layer identifier, a set identifier, and a subset identifier, and provides target resolution information for which the profile was created; and providing in their entirety over-scanned left and right eye images of the 3D images; wherein the plurality of depth points is based at least in part on a location where a human viewer is located, a display screen plane, and an infinity plane; providing the depth map in the form of a first block depth map having m rows and n columns, wherein m does not equal n; and converting the first block depth map having m rows and n columns into a reduced block depth map having p rows and q columns wherein m>p and n>q; wherein a first normalized z-axis value of the normalized z-axis values is defined as substantially equal to zero when the imaged object is perceived as to be located on the display screen plane, and a second normalized z-axis value of the normalized z-axis values is defined as approximately equal to −100 when the object is perceived to be located on the infinity plane; wherein the depth map is represented as a primary depth map and a secondary depth map, wherein the primary depth map provides depth values of an image at a pixel level or a sub-pixel level; and the secondary depth map is a block depth map that provides depth values at a greater-than-a-pixel level; wherein the converting comprises:
partitioning the first block depth map having m rows and n columns into p×q partitions, wherein each partition comprises a plurality of depth values; andusing the foremost depth values contained in each of the p×q partitions to identify corresponding depth values in the reduced block depth map having p rows and q columns, wherein the foremost value within a partition is used even when it traverses a boundary of the partition.