Depth Camera Calibration Using Natural Object Shapes

Solution Overview

Problem

Depth cameras calibrated using special test charts are inconvenient for repeated use and suffer from reduced accuracy due to thermal and mechanical stresses that change the alignment of camera components over time, limiting their usability beyond manufacturing facilities.

Innovation Solution

A system that uses natural objects with expected shapes to calibrate depth cameras, including an image receiver, object detector, edge segmenter, pose estimator, gradient generator, and parameter adjuster to reduce disparities between disparity-based and homography-based depth gradients, allowing for post-capture accuracy checks and corrections without user intervention, applicable in various environments and for different types of depth cameras.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If special test charts are used for calibration, then manufacturing precision is improved, but ease of operation deteriorates and adaptability is reduced

Engineering Contradiction:

Improvedepth camera calibration accuracyVSAvoidcalibration convenience

Core Design Contradiction:

Manufacturing precisionVSEase of operation

Solution Approach 1:

The system automatically detects natural objects with expected shapes in the scene and performs calibration without requiring manual placement of test charts or user intervention. The depth camera and reference camera autonomously identify objects, extract geometric features, and compute calibration parameters, enabling the system to calibrate itself in any environment with common objects like books, cups, or electronic devices.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The calibration system works with any natural object that has a recognizable expected shape (rectangular, circular, triangular, etc.), making it universally applicable across different environments and scenarios. Instead of requiring a specialized test chart, the system can use everyday objects found in homes, offices, or outdoor settings, thereby eliminating the need for separate calibration facilities.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Manufacturing precision

If special test charts are used for calibration, then manufacturing precision is improved, but adaptability deteriorates

Engineering Contradiction:

Improvedepth camera calibration accuracyVSAvoidenvironmental adaptability

Core Design Contradiction:

Manufacturing precisionVSAdaptability or versatility

Solution Approach 1:

The system automatically detects natural objects with expected shapes in the scene and performs calibration without requiring manual placement of test charts or user intervention. The depth camera and reference camera autonomously identify objects, extract geometric features, and compute calibration parameters, enabling the system to calibrate itself in any environment with common objects like books, cups, or electronic devices.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The calibration system works with any natural object that has a recognizable expected shape (rectangular, circular, triangular, etc.), making it universally applicable across different environments and scenarios. Instead of requiring a specialized test chart, the system can use everyday objects found in homes, offices, or outdoor settings, thereby eliminating the need for separate calibration facilities.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Measurement precision

If factory calibration is performed, then measurement precision is improved initially, but reliability deteriorates over time due to thermal and mechanical stresses

Engineering Contradiction:

Improvedepth measurement accuracyVSAvoidcalibration stability over time

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The system performs calibration periodically by automatically detecting natural objects in the scene at scheduled intervals or on-demand. This periodic recalibration compensates for drift caused by thermal expansion, mechanical stress, or component aging, ensuring that measurement precision is maintained over the long term without requiring permanent factory calibration.

Inventive Principle:

Principle #19Periodic action

Solution Approach 2:

The system continuously monitors depth measurements by comparing disparity-based depth gradients with homography-based depth gradients derived from detected natural objects. When discrepancies exceed a threshold, the system automatically adjusts calibration parameters to correct the drift, creating a closed-loop feedback mechanism that maintains reliability over time.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS10154176B1Calibrating depth cameras using natural objects with expected shapes

Publication Date: 2018.12.11 INTEL CORP

AI summary

An example apparatus for calibrating depth cameras includes an image receiver to receive an image from a depth camera. The apparatus also includes an object detector to detect a natural object with an expected shape in the image. The apparatus further includes a parameter adjuster to adjust a parameter of the depth camera to reduce a detected difference between the expected shape and a measured characteristic of the natural object.