Autostereoscopic 3D Display Region-Based View Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autostereoscopic displays face challenges in generating high-quality 3D images efficiently due to the computational demands of rendering a large number of views, leading to reduced image quality and increased resource usage, with existing approaches often introducing artifacts like ghosting and uneven viewing experiences.
Innovation Solution
An apparatus and method that generate an output 3D image by subdividing the image into regions with varying numbers of image blocks corresponding to different viewing directions, adapting the number of blocks based on local image properties such as depth and contrast, and using a combination of depth-based and non-depth-based processing to optimize computational resource usage and image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a large number of views are generated for autostereoscopic display, then the 3D effect quality is improved, but the computational complexity and resource usage increase significantly
Solution Approach 1:
The image is divided into multiple regions, and for each region, only a subset of views is generated based on local depth characteristics. This segmentation approach allows the system to process fewer views overall while maintaining 3D quality in critical areas, thereby reducing computational complexity without sacrificing manufacturing precision.
Solution Approach 2:
Different numbers of views are generated for different regions of the image based on their depth properties. Regions with significant depth variations receive more views for higher quality 3D effect, while regions with minimal depth variations use fewer views. This local quality approach optimizes the balance between 3D image quality and computational resource usage.
2Manufacturing precision
If the number of views is increased to reduce ghosting artifacts, then the perceived image quality improves, but the processing time and resource consumption increase
Solution Approach 1:
The system generates a higher number of views only in regions where ghosting artifacts are likely to occur (areas with significant depth variations), while using fewer views in regions where ghosting is less problematic. This localized approach reduces overall processing time while maintaining image quality where it matters most.
Solution Approach 2:
The system performs preliminary analysis of the image to identify regions with significant depth variations before generating views. This allows the system to pre-determine which regions require more views to prevent ghosting, optimizing the distribution of processing efforts and reducing total processing time.
3Ease of operation
If uniform number of views is generated for all regions, then the processing is simplified, but the viewing experience becomes uneven across different parts of the image
Solution Approach 1:
Instead of applying a uniform number of views across the entire image, the system adapts the number of views generated for each region based on local depth characteristics. This ensures that regions with more significant depth variations receive more views for better viewing experience, while maintaining processing simplicity through automated region-based classification.
4Measurement precision
If more image blocks are generated for each region, then the depth-based processing accuracy is improved, but the memory usage and computational load increase
Solution Approach 1:
The image is segmented into regions, and image blocks are generated selectively for each region based on depth characteristics. This segmentation allows the system to maintain high depth processing accuracy in critical regions while reducing the total number of image blocks generated, thereby optimizing memory usage and computational load.
Solution Approach 2:
The system generates the necessary number of image blocks partially - only for regions where depth-based processing accuracy is critical - rather than generating blocks for the entire image. This partial action approach maintains measurement precision where needed while reducing overall memory usage and computational requirements.
Data Source
AI summary
An autostereoscopic 3D display comprises a first unit (503) for generating an intermediate 3D image. The intermediate 3D image comprises a plurality of regions and the first unit (503) is arranged to generate a first number of image blocks of pixel values corresponding to different view directions for the region regions. The number of image blocks is different for some regions of the plurality of regions. A second unit (505) generates an output 3D image comprising a number of view images from the intermediate 3D image, where each of the view images correspond to a view direction. The display further comprises a display arrangement (301) and a driver (507) for driving the display arrangement (301) to display the output 3D image. An adaptor (509) is arranged to adapt the number of image blocks for a first region in response to a property of the intermediate 3D image or a representation of a three dimensional scene from which the first image generating unit (503) is arranged to generate the intermediate image.