Abstract
The well-being of the animals (e.g. pigs) in livestock buildings is contingent on adequate ventilation. Depending on the construction of the ventilation system draught may be introduced into the buildings. Obviously this is an unwanted effect, that might lead to decreasing growth of and increasing sickness among the livestock. Therefore it is of interest to estimate flow vector fields corresponding to the airflow introduced by a given ventilation system. By introducing particles (e.g. smoke or soapbubles) into the air-inlets of a model of a livestock building, the airflow in an laser-illuminated plane may be visualized. Based of sequences of images recorded of this plane local measurements of the velocity field are obtained by analysis of the local energy distribution, which is sampled using a set of 3-D spatio-temporal Gabor filters. We will show how physically inspired a priori distribution of the flow field may be used to integrate the local observations to a smooth field.
Original language | English |
---|---|
Title of host publication | Proceedings of the 4th Danish Conference on Pattern Recognition and Image Analysis (DANKOMB, DSAGM yearly meeting), Aug. 14-15 |
Publisher | Department of Computer Science, University of Copenhagen (DIKU) |
Publication date | 1995 |
Pages | 213-210 |
Publication status | Published - 1995 |
Event | Proceedings of the 4th Danish Conference on Pattern Recognition and Image Analysis (DANKOMB, DSAGM yearly meeting), Aug. 14-15 - Duration: 1 Jan 1995 → … |
Conference
Conference | Proceedings of the 4th Danish Conference on Pattern Recognition and Image Analysis (DANKOMB, DSAGM yearly meeting), Aug. 14-15 |
---|---|
Period | 01/01/1995 → … |
Keywords
- markov random fields
- optical flow