Infrared Pattern Detection via Difference of Gaussian Blob Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing depth sensing systems face challenges in accurately determining whether an infrared image represents a specific structured light pattern, particularly due to synchronization issues between projectors and cameras, and interference from multiple devices.

Innovation Solution

The system employs a method involving convolution with Gaussian operators to create a difference of Gaussian image, which helps identify blob regions. These regions are analyzed for geometric attributes to determine if the image represents a dot or flood pattern, facilitating the calculation of depth distances.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If infrared images are captured to determine three-dimensional shape using structured light projection, then depth sensing capability is improved, but interference from multiple devices and synchronization issues reduce measurement precision

Engineering Contradiction:
Improvedepth sensing accuracyVSAvoidinterference from multiple devices
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent segments the infrared image analysis by applying Difference of Gaussian (DoG) filtering to isolate specific spatial frequency ranges corresponding to projected pattern features. This segmentation allows the system to focus on pattern-related frequencies while filtering out interference from other devices, thereby improving depth sensing accuracy in multi-device environments

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system employs feedback mechanisms by analyzing the DoG-filtered images to detect blob regions and determine whether captured images contain the projected pattern. This feedback loop enables the system to identify and exclude images corrupted by interference or synchronization issues, maintaining measurement precision despite environmental challenges

Inventive Principle:
Principle #23Feedback

2Productivity

If multiple devices project infrared patterns simultaneously, then productivity of depth mapping is improved, but interference between devices reduces reliability of pattern detection

Engineering Contradiction:
Improvedepth mapping speedVSAvoidpattern detection accuracy
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The Difference of Gaussian filter acts as an intermediary that processes raw infrared images to extract pattern-specific features while suppressing interference from other devices. By operating in the frequency domain, this intermediary enables multiple devices to work simultaneously without compromising pattern detection reliability

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system changes parameters by adjusting the spatial frequency range analyzed through DoG filtering and by dynamically evaluating blob region characteristics. These parameter changes allow the system to distinguish between valid pattern features and interference, maintaining reliability even when multiple devices operate concurrently

Inventive Principle:
Principle #35Parameter changes

3Ease of operation

If Gaussian filtering is applied to detect blob regions, then ease of operation is improved, but computational complexity increases device complexity

Engineering Contradiction:
Improvepattern identification simplicityVSAvoidimage processing complexity
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent replaces complex mechanical or algorithmic pattern recognition systems with the mathematical Difference of Gaussian filtering approach. This substitution simplifies the overall system operation while managing computational complexity through efficient convolution operations that are well-suited for hardware implementation

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3924867B1Detection of projected infrared patterns using difference of gaussian and blob identification
Publication Date: 2025.04.09 GOOGLE LLC
  • EP3924867B1 patent drawingFigure 1
  • EP3924867B1 patent drawingFigure 2
  • EP3924867B1 patent drawingFigure 3A

AI summary

A method may include obtaining an infrared image of an object and determining a difference of Gaussian image that represents features of the infrared image that have spatial frequencies within a spatial frequency range defined by a first Gaussian operator and a second Gaussian operator. The method may also include identifying one or more blob regions within the difference of Gaussian image. Each blob region of the one or more blob regions includes a region of connected pixels in the difference of Gaussian image. The method may further include, based on identifying the one or more blob regions within the difference of Gaussian image, determining that the infrared image represents the object illuminated by a pattern projected onto the object by an infrared projector.