Self-adaptive GPS fusion positioning system and method and medium

Abstract

The invention discloses a self-adaptive GPS fusion positioning system and method and a medium. The system comprises a GNSS / INS information acquisition module, a visual information processing module and a self-adaptive fusion module. The GNSS / INS information acquisition module comprises a global satellite navigation system and an inertial navigation system; the visual information processing module is used for collecting an image of an area in front of a carrier, obtaining an area proportion of a sky area in an original image after semantic segmentation, and further obtaining pose data; the self-adaptive fusion module is used for dynamically adjusting the estimation weight of the GPS data in the fusion process of the Kalman filtering according to the area proportion of the sky area in the original image, and obtaining the filtered estimation pose; according to the method, mismatching of the sky area in visual pose estimation can be removed, and the mobile robot can obtain better pose estimation when in an urban canyon state.

Description

technical field [0001] The invention relates to the field of data fusion positioning, in particular to an adaptive GPS fusion positioning system, method and medium. Background technique [0002] In a multi-sensor information fusion system, each sensor will measure different parameters. However, the measured values ​​obtained are not only affected by the sensor's own manufacturing structure and information collection method, but also by environmental factors. GPS is a sensor that is easily affected by the environment. It cannot receive satellite signals indoors and thus cannot calculate the current position. However, when it is outdoors, the signal may still be blocked or interfered. This situation often occurs. Appears in the environment of urban canyons, because the high-rise buildings in the city will block the received signal of GPS, resulting in poor satellite tracking accuracy; the camera is a relatively stable sensor, but its positioning accuracy will be subject to som...

AI Technical Summary

Problems solved by technology

[0004] The present invention overcomes two existing problems in the prior art, one is the influence of the different accuracy of GPS data obtained in different environments on the accuracy of the entire multi-sensor information fusion system, and the other is the problem of wrong matching of specific areas in the visual pose estimation algorithm , provides a robust and adaptive GPS data fusion method, which can not only remove the mismatch of the sky region in the visual pose estimation, but also enable the mobile robot to obtain a better position when it is in the urban canyon state. pose estimation

Images