Surround-View Camera Extrinsic Calibration Using Visual Odometry

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for calibrating extrinsic parameters of surround-view cameras on vehicles are costly, require specific environments, have high computing power demands, and are not suitable for real-time online calibration.

Innovation Solution

A method and apparatus that utilize vehicle-body odometer information and visual odometer information from surround-view cameras to determine initial pose values, which are then used to calculate the extrinsic parameters of multiple cameras, forming closed-loop relationships to improve calibration efficiency and versatility.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If camera calibration method based on specific site is used, then calibration accuracy is improved, but site construction cost increases and application scenarios are limited

Engineering Contradiction:

Improvecalibration accuracyVSAvoidsite construction cost

Core Design Contradiction:

Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent replaces physical calibration templates and specific site markers with virtual markers generated through image processing. By detecting natural features in the environment and creating virtual correspondence relationships, the system achieves accurate calibration without requiring expensive physical calibration equipment or specially constructed sites.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent enables calibration to be performed in general environments without specialized markers or templates. The method can calibrate multiple cameras simultaneously using natural environmental features, making it universally applicable to various vehicle types and scenarios without requiring specific site construction.

Inventive Principle:

Principle #6Universality (Multi-functionality)

2Adaptability or versatility

If camera calibration method based on scenario map fusion is used, then calibration versatility is improved, but computing power requirement increases and calibration time consumption increases

Engineering Contradiction:

Improvecalibration versatilityVSAvoidcomputing power requirement

Core Design Contradiction:

Adaptability or versatilityVSPower

Solution Approach 1:

The patent extracts only the essential correspondence relationships between multiple cameras from complex environmental data, rather than performing full scenario map fusion. By focusing on extracting key geometric relationships and using simplified calculation models, the system achieves versatile calibration with significantly reduced computational requirements.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The calibration process is segmented into independent steps: detecting features in each camera's field of view, establishing correspondence relationships between cameras, and calculating extrinsic parameters. This segmentation allows each step to be processed efficiently without requiring excessive computing power for holistic scene understanding.

Inventive Principle:

Principle #1Segmentation

3Adaptability or versatility

If camera calibration method based on scenario map fusion is used, then calibration versatility is improved, but calibration time consumption increases

Engineering Contradiction:

Improvecalibration versatilityVSAvoidcalibration time consumption

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent performs preliminary feature detection and matching in each camera's image before calculating the final calibration parameters. By pre-processing images to identify key features and establish potential correspondences in advance, the system reduces the time required for the actual calibration computation, enabling faster and more versatile calibration.

Inventive Principle:

Principle #10Preliminary action

4Measurement precision

If visual odometer information from multiple cameras is used, then calibration accuracy is improved, but device complexity increases

Engineering Contradiction:

Improvecalibration accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges information from multiple cameras by establishing correspondence relationships between their respective fields of view. By integrating visual odometer data from adjacent cameras and using their overlapping regions to determine extrinsic parameters, the system improves calibration accuracy while managing complexity through unified mathematical models.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS12475599B2Method and apparatus for calibrating extrinsic parameters of surround-view cameras on a vehicle, vehicle and storage media

Publication Date: 2025.11.18 XUANCHENG LUXSHARE PRECISION IND CO LTD

AI summary

A method for calibrating extrinsic parameters of surround-view cameras on a vehicle includes: determining extrinsic parameters of the first camera and the second camera; determining a first pose initial value, a second pose initial value, a third pose initial value, and a fourth pose initial value; determining extrinsic parameters of the third camera and fourth camera according to the extrinsic parameters of the first camera and second camera, the first pose initial value, the second pose initial value, the third pose initial value, and the fourth pose initial value. According to the visual odometer information of the first camera and the second camera, the extrinsic parameters of the first camera and second camera are first determined, and the extrinsic parameters of the third camera and fourth camera are then determined according to the extrinsic parameter of the first camera and the extrinsic parameter of the second camera.