Video based techniques for analysing and interpreting motion information vary widely between applications. However, many of the low level vision tasks required to measure such object movement are common to many areas. Thus a recent research project has been run in the Department of Cybernetics at the University of Reading in order to design and develop a low cost, general purpose system to perform such tasks. Three main areas of interest have been identified and addressed. These include the extraction of feature positions from a real-time video signal, the tracking of these objects through the image sequence, and methods for handling the problem of occluded features. Examples of applying the resulting hardware/software system to real image sequences are also presented in the paper
Published in:
Image Processing and its Applications, 1992., International Conference on
Date of Conference: 7-9 Apr 1992