AR Pixel Blending via Alpha Matting Weighted Average

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Combining real and virtual image data in augmented reality experiences at high frame rates poses computational and power consumption challenges, particularly in head-mounted devices, where minimal data transmission and low power usage are required.

Innovation Solution

The method involves generating a combined pixel value as a weighted average of real and virtual pixel values based on alpha matting parameters, with alpha values obtained for each pixel, allowing for efficient blending of images in near real-time without significant data buffering or high power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If real and virtual image data are combined at high frame rates, then the augmented reality experience quality is improved, but power consumption and computational load increase

Engineering Contradiction:
Improveframe rateVSAvoidpower consumption
Core Design Contradiction:
ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent segments the image combination process into discrete per-pixel operations using alpha matting values. Each pixel is processed independently with a simple weighted average calculation based on pre-computed alpha values, avoiding complex global optimization and reducing computational complexity to enable high frame rates with minimal power consumption

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the computational approach by using pre-computed alpha matting parameters to guide the blending process. Instead of complex real-time image analysis, the system uses these parameters to directly compute pixel values through simple weighted averages, significantly reducing computational load while maintaining visual quality at high frame rates

Inventive Principle:
Principle #35Parameter changes

2Productivity

If real and virtual image data are combined at high frame rates, then the augmented reality experience quality is improved, but computational load increases

Engineering Contradiction:
Improveframe rateVSAvoidcomputational load
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent divides the image processing into independent per-pixel operations using alpha matting. Each pixel value is computed separately using a simple weighted average formula based on pre-computed alpha values, eliminating the need for complex computational algorithms and enabling efficient high-frame-rate processing

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary computation of alpha matting values before the actual image combination. These pre-computed parameters guide the subsequent pixel-value calculations, allowing the system to avoid complex real-time computations during the critical frame rendering phase and thereby reduce computational load

Inventive Principle:
Principle #10Preliminary action

3Manufacturing precision

If alpha matting is used for pixel blending, then image blending quality is improved, but data processing requirements increase

Engineering Contradiction:
Improveimage blending qualityVSAvoiddata processing
Core Design Contradiction:
Manufacturing precisionVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential alpha matting parameters needed for blending from the full image data. By using these extracted parameters to guide per-pixel weighted average calculations, the system achieves high-quality blending while processing minimal data, avoiding the need to analyze or transmit complete image datasets

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS11373273B2Method and device for combining real and virtual images
Publication Date: 2022.06.28 APPLE INC
  • US11373273B2 patent drawing
  • US11373273B2 patent drawing
  • US11373273B2 patent drawing

AI summary

In one implementation, a method includes obtaining, for a particular pixel of an image, a real pixel value and a virtual pixel value. The method includes obtaining, for the particular pixel of the image, a first alpha and a second alpha. The method includes generating, for the particular pixel of the image, a combined pixel value as a weighted average of the real pixel value and the virtual pixel value, the weighting being based on at least one of the first alpha and the second alpha.