Indoor Navigation by Image Recognition

Abstract

We present a prototyping mobile application for indoor navigation using image recognition to locate user's position in the indoor environment and calculate a appropriate path from the current location of user to the desired location. The user can directly follow the direction of arrow until arrived the target location, the application can able to dynamic generate a new path if the user out of path. We also make use of the sensors in the Smartphone, such as accelerometer, gyroscope and magnetometer to increase the accuracy. Our application is implemented on Android platform, but not limit on Android platform, since our navigation system is not required special features only for Android. We test our prototyping application in a shopping mall and found that our application provides better and clearer location information the paper map.