Camera Depth Image Accuracy via Phase Sensor Integration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for obtaining depth information in images, such as dual-lens cameras and structured light sensors, face challenges in accuracy and precision, especially in low-illumination scenes, leading to noise and inaccurate registration of depth information with visible light images.

Innovation Solution

A depth image obtaining method and device that processes depth information using a camera equipped with an optical lens, a phase sensor, and a visible light sensor, where the phase sensor captures infrared light to generate a phase image, and the visible light sensor captures visible light to generate a visible light image. The processor then calculates depth information from the phase image and superimposes it onto the visible light image to create a depth image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If dual-lens camera or structured light sensor is used to obtain depth information, then depth information can be acquired, but registration precision between visible light image and depth information is low and noise increases in low-illumination scenes

Engineering Contradiction:
Improvedepth information accuracyVSAvoidregistration precision
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent combines the infrared sensor and visible light sensor into a single integrated imaging device, allowing simultaneous capture of phase images and visible light images through the same optical path. This merging eliminates the registration precision issues that arise from using separate sensors and lenses, as both image types are captured from the exact same viewpoint and optical path.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent introduces a beam splitter as an intermediary optical component that divides the incoming light into infrared and visible light paths, directing them to respective sensors. This intermediary enables a single optical system to serve dual purposes, maintaining optical path consistency while separating the detection functions for different wavelength ranges.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of manufacture

If separate lenses and sensors are used for depth information and visible light image, then both can be obtained independently, but optical path alignment becomes complex and registration accuracy decreases

Engineering Contradiction:
Improveindependent sensor operationVSAvoidoptical path alignment
Core Design Contradiction:
Ease of manufactureVSDevice complexity

Solution Approach 1:

The patent merges the optical paths by using a single lens and beam splitter configuration, eliminating the need for complex alignment between separate optical systems. The integrated design reduces device complexity while maintaining the ability to independently process infrared and visible light data.

Inventive Principle:
Principle #5Merging (Combining)

3Productivity

If conventional sensor is used to obtain visible light image and depth information is superimposed through algorithm-based registration, then depth image can be generated, but registration precision is low and visible light image cannot accurately reflect depth information

Engineering Contradiction:
Improvedepth image generationVSAvoidregistration accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent performs preliminary action by capturing both the phase image and visible light image simultaneously through the same optical path at the same moment. This eliminates the need for subsequent algorithm-based registration, as the spatial correspondence is inherently established during capture, ensuring high precision depth information overlay.

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 method achieves more accurate depth information even in low-illumination scenes by utilizing the clarity of phase images, resulting in improved accuracy and precision of depth information integrated into the depth image.

Implementation Method 1

the phase sensor captures infrared light to generate a phase image

Methodology Applied
Scientific EffectInfrared radiation detection: Infrared Radiation

Implementation Method 2

The processor then calculates depth information from the phase image

Methodology Applied
Scientific EffectPhase measurement: Phase Modulation

Implementation Method 3

the visible light sensor captures visible light to generate a visible light image

Methodology Applied
Scientific EffectVisible light detection: Light

Implementation Method 4

The processor then calculates depth information from the phase image and superimposes it onto the visible light image to create a depth image

Methodology Applied
Scientific EffectImage superposition: Image Processing

Data Source

PatentEP3852362B1Method for acquiring depth image, and camera device
Publication Date: 2025.04.23 HUAWEI TECH CO LTD
  • EP3852362B1 patent drawingFigure 1~2
  • EP3852362B1 patent drawingFigure 3~4
  • EP3852362B1 patent drawingFigure 5

AI summary

This invention provides a depth image obtaining method, a terminal, and a storage medium, to process depth information to obtain more accurate depth information, and include the more accurate depth information in a depth image. The method is applied to a camera and the camera includes a visible light sensor, a phase sensor, and a processor. The visible light sensor collects a visible light image of a shooting scene. The phase sensor collects a phase image of the shooting scene. The processor obtains depth information based on the phase image, and superimposes the depth information on the visible light image to obtain a depth image.