Multidimensional Image Scaler View Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image scaling methods for multidimensional images, such as 3D-aware images, often result in the mixing of different views when upscaling or downscaling, which defeats the stereoscopic three-dimensional effect, and require substantial processing and bandwidth for generating high-resolution images.
Innovation Solution
The implementation of dimensionality-aware image scaling methods that resample pixels view by view to preserve the multidimensionality of images, allowing for efficient scaling between lower and higher resolutions while reducing the likelihood of image artifacts and enhancing perceived image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If standard image scaling methods are used on multidimensional images, then scaling between resolutions is achieved, but different views become mixed and the stereoscopic three-dimensional effect is defeated
Solution Approach 1:
The patent segments the multidimensional image into multiple distinct views before scaling operations. Each view is processed separately through the scaling pipeline, ensuring that resampling operations do not mix pixels from different views. This segmentation approach maintains view integrity while enabling efficient scaling between resolutions.
Solution Approach 2:
The patent introduces view identification metadata as an intermediary element that tags each pixel with its corresponding view information. This metadata acts as a mediator during scaling operations, guiding the resampling process to select pixels from the same view and prevent cross-view mixing while maintaining scaling efficiency.
2Manufacturing precision
If high-resolution multidimensional images are generated directly, then image quality is improved, but processing demands and bandwidth requirements increase substantially
Solution Approach 1:
The patent performs preliminary scaling operations on each individual view at lower resolutions before final composition. By preprocessing views separately at reduced resolution and then combining them with view identification metadata, the system achieves high-quality output with substantially reduced processing power consumption and bandwidth requirements compared to direct high-resolution generation.
3Productivity
If lower-resolution multidimensional images are used, then processing efficiency and bandwidth usage are improved, but image quality deteriorates
Solution Approach 1:
The patent adds the dimension of view identification metadata to the image data structure. This additional dimension enables the system to process and store images at lower resolutions while maintaining the ability to reconstruct high-quality images by properly identifying and separating views during display, thus improving processing efficiency without sacrificing image quality.
Data Source
AI summary
An electronic device uses a multidimensional (e.g., 3D) scaler to process multiple-viewing-angle (e.g., 3D-aware) images by resampling each view image and processing image data of each view image according to a view map to change resolution or improve perceived image quality. After being processed, each view image of the multiple-viewing-angle image is used to rebuild a final processed multiple-viewing-angle (e.g., 3D-aware image) with all views for displaying on the electronic device.