Vision-based Mutual Positioning Method In Unknown Indoor Environment

In our society,realizing intelligent positioning in indoor environments is important to build a smart city.Currently,mutual positioning requirements in the unknown indoor environment is growing fast.However,in such environment,we can obtain neither outdoor radio signal nor the indoor images in advance for online positioning.Therefore,how to achieve mutual positioning among users in unknown indoor environments has become a research hotspot in recent years.Aiming at the above problems,this paper proposes a vision-based mutual positioning method in unknown indoor environment.Firstly,this paper gives a brief overview of the research status of image object detection and binocular vision positioning technology at home and abroad.Secondly,this paper describes in detail the basic theories of binocular vision system modeling and feature point matching between images.Finally,this paper puts forward the following research on the key issues of the proposed method:(1)Aiming at the problem of how multiple users can find the scene that they can see together in the effective time period,this paper proposes an image information interaction and retrieval algorithm based on a shared database.When users are walking in an indoor environment,they need to take pictures of the currently observable scene at every interval,and then use the object detection network to identify all the semantic targets contained in the image and generate the corresponding semantic sequence,finally they need to upload this semantic sequence to the shared database.We can judge whether they have seen the same scene by retrieving the semantic sequences uploaded by users in the shared database.Experimental results show that this algorithm can not only effectively reduce the storage capacity of the database,but also shorten the retrieval and matching time of image information.(2)Aiming at the problem of how to carry out high-precision mutual positioning among users,this paper proposes a mutual positioning algorithm among multiple users.When users determine the scene for mutual positioning,they need to select a target in this scene as the positioning reference,and use binocular vision positioning algorithm to restore the position coordinates and turning angle information of each user relative to this target in the real world.At this time,we set up a three-dimensional coordinate system with this target as the center.After each user separately solves the position coordinates in this coordinate system,they can realize mutual positioning through the sharing of position information.Experimental results show that this algorithm can calculate the user's position coordinates and turning angle information relative to the designated target in the indoor environment with high accuracy.This paper conducts simulation experiments on the proposed mutual positioning method.The simulation results show that the proposed method can not only realize the mutual positioning among users in unknown indoor environment,but also ensure a shorter positioning time and a higher positioning accuracy,which has broad application prospect and important practical significance.