Indoor positioning method and device, electronic equipment and storage medium

Abstract

The invention discloses an indoor positioning method and device, electronic equipment and a storage medium, and relates to the technical field of positioning and the technical field of deep learning.According to the specific implementation scheme, the method comprises the following steps: extracting descriptors of structural lines in a specified image acquired by a terminal indoors; acquiring a target structural line descriptor closest to the descriptor of each structural line of the specified image from an indoor image database based on the descriptor of each structural line in the specifiedimage and the descriptor of the structural line of each image included in the pre-established indoor image database; and acquiring pose information of the terminal based on a pre-established indoor three-dimensional structure line map and each target structure line descriptor corresponding to the descriptor of each structure line of the specified image. According to the invention, the terminal can be positioned based on the structure line in the specified image acquired by the terminal indoors without being affected by texture information such as an advertising board in the image and illumination and appearance changes, higher robustness is achieved, and the stability and accuracy of indoor positioning can be effectively improved.

Description

technical field [0001] The present application relates to the field of computer technology and deep learning technology, in particular to the field of positioning technology, and in particular to an indoor positioning method, device, electronic equipment and storage medium. Background technique [0002] In indoor navigation and positioning, the current common technical solution is to firstly use image retrieval (image retrieval) technology for image retrieval to obtain a sub-map that matches the image captured by the camera. Then match the feature points in the retrieved submap, and finally use the optimization method to obtain the final camera position information. For example, when matching feature points, you can first create an offline point cloud map through structure from motion (sfm) or Simultaneous Localization and Mapping (slam) technology, and each 3D point carries coordinates information and red-green-blue (rgb) feature information. When positioning, obtain a ro...

AI Technical Summary

Problems solved by technology

[0003] However, due to the complex indoor environment such as lighting, dynamic objects, etc., the above technical solutions cannot work well

Especially in shopping malls, positioning in shopping malls based on the above features largely relies on texture-rich images such as billboards. However, billboards and other images are not static and update frequently. The stability of the technology's indoor positioning is very poor

Images