Navigation

Christoph Feldner

  • Job title: Bachelor's Thesis
  • Working group: Real-time terrain classification using inertial sensor data

Advisors
Markus Zrenner, M. Sc., Prof. Dr. Björn Eskofier

Duration
05/2018 – 10/2018

Abstract
For the recommendation of a running shoe, it is important to know the terrain that users usually run on, because the design features of running shoes differ depending on the terrain the running shoe is mainly used on. Running shoes that are usually worn in urban areas are lightweight and the soles are made of energy returning material to enable the runner to run as fast and as comfortable as possible. For these kinds of shoes, there is no need for special stability features due to the normal surfaces that the shoe is used on. In comparison, people who run on trails or uneven surfaces usually use special trail running shoes that are highly stable to prevent them from twisting their ankle while e.g. running on a forest trail with many roots on the ground. Besides, trail running shoes usually have a special profile to prevent people from falling while running on slippery surfaces [1].
For the purpose of an automated running shoe recommendation, there is a need for an objective classification of the terrain. In this work, the usage of foot mounted inertial measurement units for this task shall be evaluated. Schuldhaus et al. [2] showed that the classification of different terrains is possible. However, their approach requires a lot of computational power due to the amount of computed features. In this thesis, the aim is to develop a classification algorithm for an embedded system, which requires fewer computations due to limited energy, memory and computation resources.

References:

  1. http://www.runnersblueprint.com/running-101-5-main-running-shoes-type-select-type/; last vistited: 26.04.2018
  2. Schuldhaus, D., Kugler, P., Jensen, U., Eskofier, B., Schlarb, H., & Leible, M. (2012, November). Classification of surfaces and inclinations during outdoor running using shoe-mounted inertial sensors. In Pattern Recognition (ICPR), 2012 21st International Conference on (pp. 2258-2261). IEEE