A visual-infrared-inertial fusion positioning and mapping method

By employing a visual-infrared-inertial fusion localization and mapping method, the accuracy and stability issues of visual SLAM systems under varying lighting conditions and visual degradation environments are addressed. By using image quality assessment and adaptive weight adjustment methods, high accuracy and robustness are achieved in visually degraded environments, while maintaining high-precision localization performance in normal environments.

CN118710705BActive Publication Date: 2026-06-16HARBIN INSTITUTE OF TECHNOLOGY (SHENZHEN) (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HARBIN INSTITUTE OF TECHNOLOGY (SHENZHEN) (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)
Filing Date
2024-06-04
Publication Date
2026-06-16

AI Technical Summary

⚠Technical Problem

Existing visual SLAM systems lack positioning accuracy and stability under varying lighting conditions and visual degradation. Multi-sensor fusion systems suffer from observation errors and performance degradation in complex environments. Furthermore, existing infrared SLAM systems struggle to match the accuracy of traditional visual SLAM systems in normal environments.

⚗Method used

A vision-infrared-inertial fusion localization and mapping method is adopted. By acquiring RGB images, infrared images and IMU data, image quality is assessed, confidence is added, and factor map management module is fused in a loosely coupled manner to adaptively adjust weights. Gradient and noise indicators are used to detect environmental changes and optimize the covariance matrix to achieve adaptive weight adjustment.

🎯Benefits of technology

It improves positioning accuracy and robustness in visually degraded environments while maintaining high accuracy in normal environments, solving the performance degradation problem of visual SLAM systems under varying lighting conditions and achieving stable and accurate positioning in complex environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118710705B_ABST
    Figure CN118710705B_ABST
Patent Text Reader

Abstract

The application discloses a kind of visual-infrared-inertial fusion positioning and mapping method, comprising the following steps: S100, the image quality evaluation of respectively being carried out to the RGB image and infrared image obtained;S200, with image quality evaluation result respectively to the estimation result of VIO, the estimation result of TIO adds confidence and outputs;S300, the estimation result of VIO and the estimation result of TIO with confidence and IMU data are input into factor graph management module to be fused in loose coupling mode, obtain the pose estimation after fusion.The VIO and TIO of the present application introduce the confidence based on image quality evaluation, add image quality evaluation and realize adaptive weight adjustment factor graph management module according to confidence in the traditional loose coupling fusion framework based on factor graph, have good performance on precision and robustness in visual degradation environment, can also take into account the stability of system in visual degradation environment and precision in normal environment.
Need to check novelty before this filing date? Find Prior Art