A multi-camera calibration method

By controlling the uniformly moving target and image processing algorithms, the intrinsic and extrinsic parameters and time synchronization of multi-view cameras are automatically calculated, solving the problems of synchronization and low efficiency in multi-view camera calibration. This achieves efficient and robust multi-view camera system calibration, which is suitable for 3D reconstruction, robot navigation, and autonomous driving.

CN122244173APending Publication Date: 2026-06-19西安应用光学研究所
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
西安应用光学研究所
Filing Date
2026-03-04
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

Existing multi-camera calibration methods rely on manual operation, which is inefficient and cannot effectively solve the synchronization problem between multiple cameras, resulting in poor synchronization accuracy and sensitivity to the environment.

Method used

By controlling a target moving at a constant speed, multiple cameras synchronously acquire images and record timestamps. Image processing algorithms are used to identify target boundary information, calculate camera intrinsic and extrinsic parameters and time deviation, and perform time compensation and alignment processing to achieve time synchronization optimization of multiple cameras.

Benefits of technology

It achieves efficient automatic calibration of multi-view camera systems, improves calibration accuracy and efficiency, reduces manual intervention, and is applicable to fields such as 3D reconstruction, robot navigation, and autonomous driving.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122244173A_ABST
    Figure CN122244173A_ABST
Patent Text Reader

Abstract

This application provides a multi-view camera calibration method, comprising: controlling the trajectory and speed of a target to ensure uniform movement of the target within the field of view of the multi-view camera; synchronously acquiring target images through the multi-view camera and recording the corresponding timestamp information to obtain a time-series image; calculating the intrinsic and extrinsic parameters of the multi-view camera using a calibration algorithm based on the known correspondence between the world coordinates of the target and the pixel coordinates in the image, processing the image, and identifying the boundary information of the target motion; analyzing the motion phase information of the multi-view camera in the time-series image based on the boundary information; calculating the time deviation between the multi-view cameras based on the phase information; performing time compensation and alignment processing on the image data based on the time deviation to achieve time synchronization optimization of the multi-view camera, and finally obtaining the calibrated multi-view camera parameters. This ensures efficient collaborative operation of the multi-view camera system.
Need to check novelty before this filing date? Find Prior Art