Depth Information Storage with Multiple Values per Pixel

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional depth-determining techniques often result in inaccurate or complicated depth values due to reflections and multipath issues, leading to loss of valuable depth information when only a single depth value is stored per depth pixel, which can cause errors in image modification and depth map utilization.

Innovation Solution

Systems and techniques that store multiple depth values for each depth pixel, determining confidence intervals and statistical measures to preserve and utilize the depth information effectively, such as confidence intervals and mode values for single-peak and multi-peak distribution profiles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If multiple depth values are stored for each depth pixel, then depth information accuracy is improved, but data storage requirements and processing complexity increase

Engineering Contradiction:
Improvedepth information accuracyVSAvoiddata storage and processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments depth information into multiple discrete depth values per depth pixel, allowing the system to store and process multiple depth measurements (e.g., first depth value, second depth value, third depth value) separately. This segmentation enables preservation of multipath reflection information while maintaining manageable data structure through organized storage of multiple values per pixel.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter representation from a single depth value to multiple depth values per depth pixel. By transforming the data structure to accommodate multiple parameters (depth values, confidence intervals, distribution profiles), the system improves measurement precision while managing complexity through structured parameter organization.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If multiple depth values are stored for each depth pixel, then depth information accuracy is improved, but memory requirements increase

Engineering Contradiction:
Improvedepth information accuracyVSAvoidmemory requirements
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent segments depth information into multiple discrete depth values per depth pixel, allowing the system to store and process multiple depth measurements (e.g., first depth value, second depth value, third depth value) separately. This segmentation enables preservation of multipath reflection information while maintaining manageable data structure through organized storage of multiple values per pixel.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by storing different numbers of depth values depending on the specific depth pixel characteristics and multipath reflection conditions. Rather than uniformly storing a fixed number of values across all pixels, the system adapts the storage requirements to local scene characteristics, improving precision where needed while reducing memory usage where sufficient with fewer values.

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS20250218014A1Storing depth information
Publication Date: 2025.07.03 QUALCOMM INC
  • US20250218014A1 patent drawing
  • US20250218014A1 patent drawing
  • US20250218014A1 patent drawing

AI summary

Systems and techniques are described herein for storing depth information. For instance, a method for storing depth information is provided. The method may include obtaining depth information for depth pixels of a depth map; determining, based on the depth information, a respective plurality of depth values for each depth pixel of the depth map; and storing the respective plurality of depth values for each depth pixel of the depth map.