Vehicle Camera Calibration Using Geometric Pattern

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for determining the position and orientation of a camera in a vehicle, whether online or offline, face challenges in accuracy due to geometric assumptions and vehicle operation parameters, requiring extensive measurement efforts and being prone to errors.

Innovation Solution

A method that uses known distances between characteristic sample points on the ground, optimizing the imaging model by back-projecting pixels into the vehicle coordinate system, and determining the camera's height, roll, pitch, and yaw angles using an optimization algorithm, reducing the need for repeated measurements and improving accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If online calibration methods are used during vehicle operation, then calibration can be performed without stopping the vehicle, but accuracy is limited due to geometric assumptions and vehicle operation parameters

Engineering Contradiction:
Improvecalibration convenienceVSAvoidcamera position accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent applies preliminary action by pre-defining a geometric pattern with known characteristic points that can be used for calibration. The pattern is designed in advance with precise geometric relationships, allowing the calibration process to benefit from pre-established accuracy references rather than relying on assumptions during operation.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The geometric pattern serves as an intermediary object between the camera and the calibration process. Instead of directly calibrating using vehicle operation parameters or environmental assumptions, the pattern acts as a mediator with known geometric properties that bridge the gap between the camera and the calibration algorithm, enabling accurate determination of camera position and orientation.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If known offline calibration methods are used with geometric patterns, then accuracy can be improved, but the position of characteristic points must be precisely known requiring considerable time and equipment

Engineering Contradiction:
Improvecamera position accuracyVSAvoidcalibration time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The geometric pattern is designed in advance with known characteristic points and precise geometric relationships. This preliminary preparation of the calibration pattern eliminates the need for time-consuming measurements during the calibration process itself, as the geometric properties are already established and known.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

Instead of requiring physical measurement of characteristic points during calibration, the patent uses a geometric pattern where the positions are defined by their geometric relationships to each other. The calibration process copies or references these known geometric relationships rather than measuring absolute positions, significantly reducing the time and equipment needed.

Inventive Principle:
Principle #26Copying

3Measurement precision

If three coordinates of characteristic points are determined precisely in the vehicle coordinate system, then calibration accuracy is improved, but measurement effort and susceptibility to errors increase

Engineering Contradiction:
Improvecharacteristic points position accuracyVSAvoidmeasurement equipment and process complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the parameter representation from absolute three-dimensional coordinates in the vehicle coordinate system to relative two-dimensional distances between characteristic points. This parameter transformation simplifies the measurement requirements, as only inter-point distances need to be known, not the absolute positions in 3D space, thereby reducing measurement complexity and error susceptibility.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent extracts only the essential geometric information needed for calibration - the distances between characteristic points - while eliminating the need to determine full three-dimensional coordinates. By taking out only the necessary parameters (inter-point distances) and discarding redundant information (absolute positions), the measurement process becomes simpler and less error-prone.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP2166510B1Method for calculating the position and orientation of a camera in a vehicle
Publication Date: 2018.03.28 DELPHI TECHNOLOGIES INC
  • EP2166510B1 patent drawingFigure 1~2

AI summary

The method involves arranging a camera (13) at a distance (19) from a geometrical sample (15), such that three characteristic sample points in a sight area of the camera are not in a straight line. An image of the geometrical sample is taken by the camera, where a height, a rolling angle and pitch angle of the camera are determined relative to defined coordinate system of a vehicle (11). Independent claims are included for the following: (1) a method for calibrating a camera of a vehicle support system of a motor vehicle; and (2) an arrangement for determining a position and orientation of a camera in a vehicle, particularly for use in a vehicle support system.