A Novel Indoor Localization Method Using Image and Virtual Grid and Its System

摘要

The levels of social informatization and digitization are increasing day by day, so there is an increasing demand of indoor accurate positioning. After studying the existing positioning technologies, the paper puts forward a novel indoor localization method using image and virtual grid and its system. Firstly, indoor floor is divided into several virtual grids, where scene images are shot in four directions, and SURF algorithm is used to extract feature points, establishing image feature database. Secondly, when shooting a test image, the search space is got according to the direction information of its electronic compass, where the best matching image is searched by using binary search and its maximum probability grid is determined, to realize user coarse positioning. Finally, different weights are set to the centre coordinates of the maximum probability grid and its 8 neighborhoods according to the number of matching points, and their joint calculation is applied to complete user accurate positioning. The experimental results show that the study has accurate positioning, which error is not more than 30 cm, and does not drift with time. Meanwhile, its system does not require other hardware installed on mobile phone, only uses the camera of phone configuration to complete positioning, having advantages of low cost, high reliability and good generality.