Kinect sensor image processing techniques for the automatic motion of a wheelchair

Dissertation Information

  • Fouskas Konstantinos

Dissertation Description

Mobility in everyday life seems to be one of the most decisive factors, if not the most crucial, for the quality of life of every human being. However, unfortunately, the possibility that many of us take for granted, is not given by default to a large number of our fellow citizens with disability. A restricted mobility can result in poor quality of life and other problems. This thesis is looking to help users of electric wheelchairs, with regard to mobility problems in everyday life.

It focuses on an obstacle detection system, which will be based on: 1) image processing algorithms, which will be collected by a sensor, based on the RGB-D camera (Microsoft Kinect XBOX 360), which is placed in a wheelchair’s arm and is considered part of driving system, 2) the processing of the image and 3) the manner in which the obstacles are detected in a static or dynamic environment.

The above obstacle detection subsystem, will be based on the processing of point clouds, thanks to the help of Point Cloud Library of the programming language C++.

This subsystem is one of the major subsystems of the overall system, whose purpose is to support a semi – automatic movement of the electric wheelchair. However, as is readily appreciated, the use of such drive system requires the full attention and general capacity of the user so as to navigate the electric wheelchair in a static or dynamic environment.