Vehicle Camera Calibration at Standstill Using Image Flux Vectors

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing vehicle surround view systems require online calibration during driving operations, which may not accurately represent vehicle surroundings, especially when the vehicle is loaded, leading to incorrect distance estimation and unsuitable surround view representation during parking or driving off.

Innovation Solution

A method and device that calibrate vehicle cameras at a standstill by reading and comparing camera images to set calibration parameters, using flux vectors and reference vectors, and accounting for camera movements caused by door and mirror adjustments, ensuring correct calibration before driving.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If online calibration is performed during driving operation, then calibration correction can be achieved, but the calibration accuracy deteriorates due to vehicle movement and loading changes

Engineering Contradiction:
Improvecalibration accuracyVSAvoidcalibration timing
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs calibration at vehicle standstill before driving operation begins, rather than during driving. This preliminary calibration ensures accurate distance estimation and surround view representation are established before the vehicle moves, avoiding the inaccuracies that occur when calibration is attempted during motion or after loading changes have already affected camera positions

Inventive Principle:
Principle #10Preliminary action

2Adaptability or versatility

If calibration is performed during driving operation, then calibration updates can be made, but the representation of vehicle surroundings becomes inaccurate especially when vehicle is loaded

Engineering Contradiction:
Improvecalibration update capabilityVSAvoiddistance estimation accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The system performs calibration before driving operation and before loading changes can affect the vehicle. By establishing the calibration baseline when the vehicle is at standstill and presumably in a known loading state, the system prevents calibration errors from occurring in the first place, rather than attempting to correct them after loading has distorted camera positions

Inventive Principle:
Principle #9Preliminary anti-action

3Ease of manufacture

If static calibration rule is stored from design data, then calibration is simple to implement, but calibration errors occur due to manufacturing tolerances and loading changes

Engineering Contradiction:
Improvecalibration implementation simplicityVSAvoidcalibration accuracy
Core Design Contradiction:
Ease of manufactureVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where the system actively performs calibration routines at vehicle standstill, using actual camera images and detected features to determine calibration parameters. This feedback loop allows the system to compensate for manufacturing tolerances and loading changes by measuring the actual camera positions and adjusting calibration accordingly, rather than relying solely on pre-stored static calibration data

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS10438374B2Method and device for calibrating a vehicle camera of a vehicle
Publication Date: 2019.10.08 ROBERT BOSCH GMBH
  • US10438374B2 patent drawing
  • US10438374B2 patent drawing
  • US10438374B2 patent drawing

AI summary

A method for calibrating a vehicle camera of a vehicle includes at least one task of reading in and one task of setting. In the reading in, at least one first camera image and one second camera image are read in, which represent images recorded by at least the vehicle camera during a camera movement of the vehicle camera at a standstill of the vehicle. In the setting, a calibration parameter for calibrating at least the vehicle camera is set, using the first camera image and the second camera image.