Multi-view 3D Display Parallax Expansion via Stereo Interpolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D display technologies require special glasses or headgear and have limited viewing angles, restricting the number of viewers and freedom of movement, as they rely on stereo views that only provide a narrow range of parallax, making them less acceptable to the general public.
Innovation Solution
A method to convert stereo views into multiple views using software that generates up to 100 or more views by interpolating and synthesizing images, allowing 3D imagery to be displayed from various angles without the need for glasses or headgear, using a barrier grid or lenticular lens array on display devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If stereo views are used for 3D display, then 3D imaging is achieved, but viewing angle is limited and requires special glasses
Solution Approach 1:
The patent segments the 3D image into multiple views (typically 7 views including leftmost, rightmost, and five intermediate views) that can be displayed simultaneously on the screen. Each view corresponds to a different horizontal viewing angle, allowing viewers to see the appropriate view based on their position without requiring special glasses
Solution Approach 2:
The patent adds the dimension of multiple viewing angles by generating images from different horizontal positions. Instead of providing a single stereo pair, the system creates a sequence of views across the horizontal dimension, enabling viewers to access different perspectives by moving laterally across the viewing zone
2Adaptability or versatility
If stereo views with limited parallax are used, then 3D effect is achieved, but viewing freedom and mobility are restricted
Solution Approach 1:
The system pre-calculates and generates multiple views covering a wide range of horizontal angles before the viewer arrives. By anticipating all possible viewing positions and preparing images for each angle in advance, the system enables viewers to move freely without losing the 3D effect or requiring real-time image generation
Solution Approach 2:
The patent changes the parallax parameter by generating views with progressively increasing horizontal angles from the leftmost to rightmost positions. This creates a continuous range of parallax values that adapts to the viewer's lateral position, maintaining the 3D effect across a wide viewing zone
3Area of stationary object
If multiple views are generated from stereo images, then viewing angle is widened, but image processing complexity increases
Solution Approach 1:
The system creates copies of the original stereo image pair, transforming each into a sequence of views. By copying and adapting the image processing pipeline for each view rather than attempting complex real-time rendering, the system efficiently generates multiple views from static source material
Solution Approach 2:
The patent uses intermediate views as mediators between the original stereo pair and the final multi-view output. The five intermediate views serve as stepping stones that bridge the leftmost and rightmost extremes, allowing the system to distribute processing across multiple simpler transformations rather than attempting to generate all views simultaneously
Data Source
AI summary
A method for generating additional views from a stereo image defined by a left eye image and a right eye image. The method includes receiving as input at least one stereo image. The method includes, for each of the stereo images, generating a plurality of additional images. The method includes interlacing the additional images for each of the stereo images to generate three dimensional (3D) content made up of multiple views of the scenes presented by each of the stereo images. The interlacing may be performed such that the generated 3D content is displayable on a 3D display device including a barrier grid or a lenticular lens array on the monitor screen. The additional images may include 12 to 40 or more frames providing views of the one or more scenes from differing viewing angles than provided by the left and right cameras used to generate the original stereo image.