3D Display Subtitle Rendering Depth Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autostereoscopic and stereo 3D displays cause viewer fatigue due to accommodation-convergence mismatch, blurring of text, and geometrical differences between left and right eye views, especially when subtitles are not optimally positioned, leading to impaired readability and visual strain.
Innovation Solution
A method to improve subtitle positioning on 3D displays by controlling disparity, maintaining constant depth, and adjusting 3D parameters to reduce ghosting and blurring, ensuring subtitles remain in front of video objects and are positioned within a limited depth range, using metadata to inform the display about subtitle location and adjusting lens configurations for clearer rendering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If subtitles are displayed on autostereoscopic 3D displays with standard rendering, then the display can show multiple views to create 3D impression, but the text becomes blurred due to crosstalk and resolution sacrifice, reducing readability
Solution Approach 1:
The patent applies local quality by treating the subtitle region differently from the rest of the image. The rendering device identifies subtitle locations and applies specific rendering parameters (such as reduced disparity, limited depth range, or disabled view separation) only to those regions, while maintaining full 3D rendering capabilities for other areas of the display.
2Adaptability or versatility
If disparity is increased to enhance 3D effect, then the 3D impression is improved, but visual fatigue and headaches increase due to accommodation-convergence mismatch
Solution Approach 1:
The patent changes rendering parameters dynamically based on content type and viewer settings. By adjusting disparity values, depth ranges, and convergence points according to what is displayed (e.g., reducing disparity for subtitles while maintaining it for 3D objects), the system optimizes the balance between 3D effect strength and viewer comfort, preventing accommodation-convergence mismatch.
3Manufacturing precision
If text is positioned at depth neutral location, then readability is improved, but the text may be occluded by video objects with greater depth
Solution Approach 1:
The patent applies preliminary action by pre-positioning subtitles at depth-neutral locations and pre-calculating potential occlusion issues. The system proactively adjusts rendering parameters before occlusion occurs, such as modifying the depth buffer handling or applying depth offsetting techniques to ensure subtitle pixels remain visible even when video objects with greater depth are present in the same spatial region.
4Adaptability or versatility
If resolution is sacrificed to generate multiple views for autostereoscopic display, then 3D multi-view capability is achieved, but text bandwidth decreases below two cycles per character, impairing readability
Solution Approach 1:
The patent applies segmentation by separating the display area into distinct regions: subtitle regions and video content regions. For subtitle regions, the system uses dedicated rendering paths that maintain higher resolution and do not participate in the view-generation downsampling process. This allows multiple views to be generated for the video content while preserving full resolution for text overlay regions.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method of creating a three-dimensional image signal comprises receiving a first image component, receiving a second component for creating a three-dimensional image in combination with the first image component, receiving a text component for including in the three-dimensional image, receiving a data component comprising location information describing the location of the text component within the three-dimensional image, and creating a three-dimensional image signal comprising the first image component, the second component, the text component, and the data component. The signal is rendered by rendering a three-dimensional image from the first image component and the second component, the rendering including rendering the text component in the three-dimensional image, the rendering of the text component including adjusting three-dimensional parameters of the three-dimensional image in the location of the rendered text component.