Multi-pose pedestrian indoor positioning method fused with visual information

By establishing a mobile phone holding posture recognition model and a multi-posture PDR method that fuses visual information, the heading angle is identified and compensated. Combined with the EKF algorithm to correct the cumulative error, the problem of low positioning accuracy and cumulative error in the existing technology is solved, and high-precision indoor pedestrian positioning is achieved.

CN116793358BActive Publication Date: 2026-06-23CHONGQING UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CHONGQING UNIV
Filing Date
2023-06-13
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

Existing indoor pedestrian positioning methods have low positioning accuracy, PDR algorithms are difficult to adapt to changes in pedestrian posture, and the cumulative errors of inertial devices are difficult to eliminate autonomously.

Method used

By establishing a mobile phone holding posture recognition model, and using a multi-pose PDR method based on visual information fusion to identify and compensate for heading angles, a multi-pose PDR is constructed. Combined with the EKF algorithm, the accumulated error is corrected, thereby improving positioning accuracy.

Benefits of technology

It effectively improves the robustness of the PDR algorithm in attitude change scenarios, adapts to pedestrian attitude changes and corrects accumulated errors in real time, enhances positioning accuracy and robustness, and reduces the burden of the positioning system on pedestrians.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116793358B_ABST
    Figure CN116793358B_ABST
Patent Text Reader

Abstract

The application provides a multi-pose pedestrian indoor positioning method fused with visual information, comprising the following steps: S1. establishing a mobile phone holding pose recognition model; S2. determining the mobile phone holding pose by using the mobile phone holding pose recognition model, and then compensating the heading angle of the PDR algorithm; S3. introducing a judgment algorithm to remove the meaningless compensation heading angle in the pose conversion process and redetermine the compensation heading angle; S4. based on the compensation heading angle redetermined in step S3, constructing a multi-pose PDR and outputting a positioning result; and S5. designing a PDR error correction model based on visual position recognition, eliminating cumulative errors and further improving the positioning accuracy of the multi-pose PDR. The multi-pose pedestrian indoor positioning method fused with visual information can adapt to the pose transformation of pedestrians and can correct the cumulative errors in real time, can enhance the precision and robustness of the pedestrian indoor positioning in the case of reducing the burden of the positioning system on the pedestrians and compressing the positioning cost to the greatest extent.
Need to check novelty before this filing date? Find Prior Art