A Mobile Outdoor Augmented Reality Method Combining Deep Learning Object Detection and Spatial Relationships for Geovisualization.

Jinmeng Rao,Yanjun Qiao,Qingyun Du,Junxing Wang,Fu Ren

doi:10.3390/s17091951

Abstract

The purpose of this study was to develop a robust, fast and markerless mobile augmented reality method for registration, geovisualization and interaction in uncontrolled outdoor environments. We propose a lightweight deep-learning-based object detection approach for mobile or embedded devices; the vision-based detection results of this approach are combined with spatial relationships by means of the host device’s built-in Global Positioning System receiver, Inertial Measurement Unit and magnetometer. Virtual objects generated based on geospatial information are precisely registered in the real world, and an interaction method based on touch gestures is implemented. The entire method is independent of the network to ensure robustness to poor signal conditions. A prototype system was developed and tested on the Wuhan University campus to evaluate the method and validate its results. The findings demonstrate that our method achieves a high detection accuracy, stable geovisualization results and interaction.

Highlights

Geovisualization, or geographic visualization, is an efficient way to describe the real geographic world through visual means, thereby making complex geographic data and information intuitive and easy to understand
Many traditional Augmented Reality (AR) methods have several drawbacks when applied in various outdoor environments: fiducial-marker-based methods are inappropriate for use in uncontrolled environments, whereas sensor-based methods can be overly sensitive to certain variables in outdoor environments, such as magnetic fields, which can cause errors and even failures
We proposed a robust, markerless and near-real-time mobile outdoor AR method for geovisualization

Summary

Introduction

Geovisualization, or geographic visualization, is an efficient way to describe the real geographic world through visual means, thereby making complex geographic data and information intuitive and easy to understand. An appropriate geovisualization method can provide prompt insight and understanding to support real-world knowledge construction and decision-making [1]. Traditional geovisualization usually refers to 2D/3D cartographic visualization, which, to some degree, is isolated from the real world because it involves creating another “world” (such as a map or a virtual environment) to describe the real world, and this isolation may result in improper spatial cognition or may even produce incorrect information. From multiple perspectives (from different directions) as well as under many different conditions to allow these geographic objects to be detected in images at any possible size, from various possible. VOC2007 geographic (the entire allow these geographic informat imagesfor at the anyselected possible size, fromobjects various possible process is illustrated in 10)

Objectives

Results

Discussion

Conclusion