Camera Calibration via Ad Hoc Feature Homography

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing camera systems in consumer devices, such as smartphones and tablets, face challenges in maintaining calibration due to factors like shifting and noise, which affect image quality and require on-the-fly calibration to ensure resilient performance.

Innovation Solution

An ad hoc calibration process is implemented using a computing device with modules like feature detection, homography calculation, and rectification, leveraging factory calibration data and real-world images to refine camera alignment and estimate scene depth, thereby correcting image alignment and enhancing image quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If camera systems are calibrated during manufacturing, then initial calibration is achieved, but calibration drift occurs over time due to shifting and noise

Engineering Contradiction:
Improveinitial calibration accuracyVSAvoidcalibration stability over time
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent implements dynamic calibration by continuously updating camera parameters during operation rather than relying solely on static manufacturing calibration. The system captures images, detects features, and recalibrates camera intrinsic and extrinsic parameters in real-time to compensate for drift caused by shifting and noise, thereby maintaining reliability over time while preserving initial manufacturing precision.

Inventive Principle:
Principle #15Dynamics

2Reliability

If ad hoc calibration is performed as images are captured, then calibration stability is maintained, but processing time and computational complexity increase

Engineering Contradiction:
Improvecalibration stability over timeVSAvoidcalibration processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies partial calibration by selectively updating only certain camera parameters (intrinsic or extrinsic) based on detected feature quality and calibration needs, rather than performing full recalibration on every image. This approach maintains calibration stability while reducing processing time by avoiding unnecessary computational overhead.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system implements periodic calibration by performing full calibration at intervals and using lighter-weight adjustments between periods. This balances the need for calibration stability with processing time constraints by not continuously executing computationally intensive calibration routines on every captured image.

Inventive Principle:
Principle #19Periodic action

3Measurement precision

If feature detection and homography calculation are used for calibration, then image rectification accuracy improves, but device complexity increases

Engineering Contradiction:
Improveimage rectification accuracyVSAvoidcalibration system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements self-service calibration by using the camera system itself to detect features and calculate homography transformations without requiring external calibration equipment or complex additional hardware. The system uses naturally captured images and their features to perform calibration, thereby improving rectification accuracy while minimizing the increase in device complexity.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3189658B1Camera calibration
Publication Date: 2019.07.24 INTEL CORP
  • EP3189658B1 patent drawingFigure 1
  • EP3189658B1 patent drawingFigure 2
  • EP3189658B1 patent drawingFigure 3

AI summary

Techniques for image calibration are described herein. The techniques may include detecting features on a set of images, describing features on the set of images, determining a match between features of the image sets, determining a shift on the matched features based on camera positions associated with the matched features, determining a first homography between the camera positions and the determined shift, and determining a second homography based on a re-projection of three-dimensional features back to the cameras.