GPS technology is used by most navigation systems today to track the user’s movements. Among GPS’s disadvantages are inconsistent reception, the ability of hostiles to intercept signals, inaccuracy of tracking at very short distances, and the inability to classify movement types and steps taken.
Tracking and identifying steps has many applications, such as augmented reality museum tours, fitness apps, navigation in public spaces, and more. Our goal was to create a system that could evaluate the route of the user without using GPS and to classify the steps taken by the user. The system uses the accelerometer, the gyroscope, and the device orientation of the smartphone to accomplish this.
Several tasks were performed to reach the project’s goal: developing an Android app to collect information from the smartphone, designing a neural network model to classify steps and their length, developing a system to generate a database for network’s training, and constructing a comprehensive system to display the user’s route and types of steps taken.