3D Imaging Depth Analysis via Pattern Modulation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing 3D imaging techniques using active stereo depth estimation face challenges in accurately estimating depth within homogeneous interior regions of objects due to the visibility of projected patterns and the need for additional infrared image sensors, which introduces errors and requires complex processing.

Innovation Solution

A method for 3D imaging that involves controlling a light projector to illuminate a scene with multiple light projections having different patterns, acquiring corresponding image frames from cameras, and combining these images to produce a combined image frame where the projection patterns are minimized, allowing for accurate depth analysis using visible light without introducing errors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If visible light projection patterns are used for active stereo depth estimation, then depth analysis accuracy in homogeneous regions is improved, but the projection patterns become visible in the output images affecting user experience

Engineering Contradiction:
Improvedepth analysis accuracyVSAvoidvisibility of projection patterns
Core Design Contradiction:
Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The system uses periodic modulation of the projection pattern at frequencies above human visual perception thresholds. The projection pattern is switched between different states (e.g., different phases or frequencies) in temporal sequence, allowing depth information to be extracted while the patterns remain imperceptible to human observers. This resolves the contradiction by making the harmful visible patterns imperceptible through temporal periodicity.

Inventive Principle:
Principle #19Periodic action

Solution Approach 2:

The system employs projection patterns in the infrared spectrum rather than visible light. By changing the wavelength/frequency of the projection into the infrared range, the patterns provide sufficient contrast for depth analysis while being invisible to human eyes. This directly resolves the contradiction by using color/wavelength changes to separate the functional visibility (for depth analysis) from perceptual visibility (for users).

Inventive Principle:
Principle #32Color changes

2Object-affected harmful factors

If infrared light is used for projection to avoid visible patterns, then pattern visibility is reduced, but additional infrared image sensors are required increasing system complexity

Engineering Contradiction:
Improvevisibility of projection patternsVSAvoidnumber of image sensors
Core Design Contradiction:
Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The system uses a single RGB camera that performs dual functions: capturing visible light for color information and capturing modulated infrared projection patterns for depth information through computational techniques. The same sensor hardware is used for both color and depth acquisition, eliminating the need for separate infrared sensors. This resolves the contradiction by making one device perform multiple functions.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system creates a computational copy of the depth information by processing the RGB camera output through techniques such as temporal differencing or frequency domain analysis. Instead of using dedicated infrared sensors, the system generates depth maps by computationally extracting depth-related signals from the standard RGB camera data, effectively copying depth information from visible light captures through mathematical transformation.

Inventive Principle:
Principle #26Copying

3Measurement precision

If multiple light projections with different patterns are used for depth analysis, then depth estimation accuracy is improved, but the processing complexity and time increase

Engineering Contradiction:
Improvedepth estimation accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary action by capturing multiple sequentially projected patterns (such as phase-shifted sinusoidal patterns or coded aperture sequences) in rapid succession before combining them. The temporal separation of pattern projections allows the system to pre-capture all necessary data for depth calculation, which is then processed together to achieve high accuracy. This resolves the contradiction by using temporal preliminary actions to enable accurate depth estimation without excessive processing delays.

Inventive Principle:
Principle #10Preliminary action

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach enables accurate depth estimation within homogeneous regions by reducing the visibility of projection patterns in the output images, improving the reliability of depth analysis, and eliminating the need for additional infrared sensors, thus enhancing the efficiency and accuracy of 3D imaging systems.

Implementation Method 1

Active stereo techniques address this problem by projecting patterned light onto the scene being imaged, thus providing 'texture' to the interiors of surfaces

Methodology Applied
Scientific EffectLight reflection: Reflection

Implementation Method 2

controlling one or more cameras to each acquire a first image frame of the scene illuminated by the first light projection

Methodology Applied
Scientific EffectLight detection: Photoelectric Effect

Data Source

PatentEP4563938A1System and method for 3D imaging
Publication Date: 2025.06.04 KONINKLIJKE PHILIPS NV
  • EP4563938A1 patent drawingFigure 1~2
  • EP4563938A1 patent drawingFigure 3~5
  • EP4563938A1 patent drawingFigure 6

AI summary

A method and system for 3D imaging. A light projector illuminates a scene with a first light pattern, and an image of the scene illuminated with the first light pattern is acquired. The light projector then illuminates the scene with one or more further light patterns, and, for each further light pattern, an image of the scene illuminated by the further light pattern is acquired. Depth analysis is performed on at least one of the images. The images are combined to produce a combined image, in which the first light pattern and each of the one or more further patterns are less perceptible.