Research Group Efficient Embedded Systems THA-Logo
TH Augsburg / Faculty of Computer Science / Research Group EES german
Home
Team
Job vacancies
Open
degree theses
Lab
Research
PicoNut
THANNA
ParaNut
ASTERICS
ChASA
Triokulus
e4rat
HiCoVec
Teaching
Publications
Imprint

Spotty - An Autonomous Service Robot

Project Work (IMSc1) in SS 2012 and WS 2012/13

Definition of the project:

The goal of this project work is to develop an autonomous mobile robot that can orient itself on the basis of optical characteristics in the room.

For example, the robot is supposed to detect objects and to transport them to a particular location. The physical platform for the system is a vacuum cleaner robot which is already available. With the support of additional embedded hardware (Atom or ARM platform) it should be able to recognize and to interact with certain objects (transport). Other objects which are fixed in space should be utilized for guidance.

The biggest challenge is the recognition of the environment and the determination of the robots own position. Currently available vacuum cleaner robots will need, for example, specific (radio) markers, which must be placed in the room. In this project work the navigation should solely be based on natural optical characteristics of the environment. This way, the space in which the robot moves, does not need to be specially prepared.

For the solution of the task, the experiences of the BMBF project "Triokulus" can be utilized. Over the past few years, innovative algorithms and hardware modules for the recognition of natural objects have been developed and published internationally. According to success, the result of the project work could also lead to an international publication.

Requirements:

  • Interest in computer vision
  • Good knowledge of C/C++
  • Interest in the development of embedded systems
  • Experience with Linux is an advantage

Material:

Slides Preliminary discussion (Single slides)
Literature/Code Diploma Thesis by Konstantin Otto
OpenCV Demo - SURF
OpenCV Demo - Image Processing
SVN-Server of the Triokulus-Project
17.4.2024 - Michael Schäferling - Technische Hochschule Augsburg