Neural Text Reconstruction for Legible XR Image Output
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
XR headsets face limitations in displaying textual content at high resolution due to constraints in optics, hardware capabilities, frame rates, and lighting variations, leading to less than perfect resolution and clarity.
Innovation Solution
An apparatus and method using an image sensor and neural network to detect and enhance textual features in XR headsets by deducing multiple glyphs that form legible words or abbreviations, even in low resolution or noisy conditions, through semantic analysis and domain-specific training.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high quality hardware and software are used to display textual content in XR headsets, then resolution and clarity are improved, but cost increases and feasibility decreases
Solution Approach 1:
The patent changes the parameter of text rendering from native low-resolution display to super-resolved high-resolution reconstruction. The neural network processes low-resolution input text and transforms it into high-resolution output by learning the mapping between downsampled and original high-resolution text images, achieving high visual quality without requiring high-resolution display hardware
Solution Approach 2:
The patent creates a virtual copy of high-resolution text through neural network synthesis. Instead of displaying actual high-resolution text directly on the limited hardware, the system generates a synthetic high-resolution version by upscaling and refining the low-resolution input, effectively copying the visual appearance of high-quality text without the hardware cost
2Device complexity
If native low resolution display is used in XR headsets, then hardware complexity is reduced, but textual content clarity deteriorates
Solution Approach 1:
The patent introduces a neural network processing system as an intermediary between the low-resolution display hardware and the user's visual perception. This intermediary software layer performs super-resolution reconstruction, transforming the low-resolution native display output into high-clarity visual text, thereby decoupling hardware simplicity from visual quality
3Use of energy by moving object
If text is displayed at native resolution in XR headsets, then power consumption is reduced, but reading comfort and user experience deteriorate
Solution Approach 1:
The patent applies partial super-resolution processing only to textual regions rather than the entire image. The system identifies text areas and applies super-resolution enhancement selectively to those regions, consuming minimal additional power while significantly improving reading comfort for the most critical visual elements
Data Source
AI summary
Disclosed is an apparatus including an image sensor to capture an input image with one or more textual features. The one or more textual features are present in an unreadable or distorted form. Further, the apparatus includes a processor configured to detect the one or more textual features of the input image and execute a neural network to concurrently deduce a plurality of glyphs that form one or more relevant words or abbreviations based on the detected one or more textual features of the input image. Further, the processor is configured to generate an output image with enhanced one or more textual features in a legible form based on the deduced plurality of glyphs.