Azimuth tracking for a rolling shutter camera

By combining inertial sensor data and visual information, the camera orientation tracking method for rolling shutter cameras was improved. The camera's velocity changes during the readout time were dynamically modeled, which solved the problem of inaccurate orientation estimation of rolling shutter cameras in augmented reality systems, and improved the accuracy of image tracking and user experience.

CN116348916BActive Publication Date: 2026-06-19SNAP INC
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SNAP INC
Filing Date
2021-09-22
Publication Date
2026-06-19

AI Technical Summary

Technical Problem

In augmented reality systems, when using rolling shutter cameras, existing technologies struggle to accurately estimate camera orientation, especially when the camera's speed varies inconsistently during readout time, leading to tracking failures or distortions.

Method used

By combining inertial sensor data and visual information, the camera orientation tracking method is improved. The camera's velocity change during the readout time is dynamically modeled, the camera orientation of each scan line is estimated using accumulated IMU data, and the image information is combined for refinement.

Benefits of technology

It improves the accuracy of camera orientation estimation, reduces image distortion, ensures seamless overlay of the virtual and real worlds in augmented reality systems, and provides a better user experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116348916B_ABST
    Figure CN116348916B_ABST
Patent Text Reader

Abstract

This disclosure provides a method and apparatus for tracking the orientation of a rolling shutter camera in an augmented reality (AR) system. The method and apparatus use camera information and inertial sensor readings from an inertial measurement unit (IMU) to estimate the camera's orientation at a reference line. The inertial sensor data is then used to calculate the relative orientation change at the scan line. The estimated reference orientation of the camera is then further refined based on visual information from the camera, the relative orientation change, and an optimized reference line orientation from a previous image. The estimation of the scan line orientation can then be updated using the relative orientation change obtained in an earlier step.
Need to check novelty before this filing date? Find Prior Art