One way to tackle the problem of dust concentrations in pig housing is avoiding the peak concentrations that are caused by periods of agile movements of the pigs. However, to control the activity of the pigs there is still a lack of sensors to measure their activity level continuously in an on-line way. The purpose of this project was to develop a method for measuring the activity level of pigs in a pen in real time and identify an on-line behavior model to detect and predict occurrences of aggressiveness of pigs. For the experiments, 2 adjacent pig pens were used, each 6.9 by 2.6 meters in size and each populated with about 15 pigs. The equipment used for activity measurement consisted of an infrared-sensitive CCD camera that was mounted to the roof, 5 meters above the floor of the pen and connected to a PC with built in frame grabber. The camera was in a protective housing to shield it from dust and moisture. The lens was pointed downwards to get a top view of the pen. Images were captured with a resolution of 768 by 586 pixels and a 1 Hz frame rate. Software was developed to measure the activity level of animals, visible in the camera image, in real time and in practical conditions, i.e. a pigpen. Before the experiments, the two pens visible in the camera images were each divided in two zones, covering the left and right hand side, respectively. Every second, the algorithm logged the camera image and the activity index for each zone, defined as the fraction of the floor space in the pen that was covered or uncovered by pigs in the camera image during a one second period, i.e., the activity index is the fraction of floor space that is 'moving'. Finally the recorded video images were visually labeled in order to score animals' behavior and find a relation with the automatically measured activity index. From the analysis of the automatically measured group activity index compared to the manual labeling a relation exists between the activity index and the behavior types 'no activity', 'nuzzling' and feeding \ However, behavior related to aggressiveness of individual pigs ('fighting' and 'biting') can not be detected from instantaneous measurements of the group activity index.

Real-time measurement of pig activity in practical conditions / T. Leroy, F. Borgonovo, A. Costa, J.M. Aerts, M. Guarino, D. Berckmans - In: Livestock Environment VIII : ASABE 8. International Symposium / [a cura di] D. De Moura D, R.S. Gates. - [s.l] : ASABE, 2008 Aug. - pp. 12-12 (( Intervento presentato al 8. convegno International Livestock Symposium tenutosi a Iguassu Falls, Brazil nel 2008.

Real-time measurement of pig activity in practical conditions

F. Borgonovo;A. Costa;M. Guarino;
2008

Abstract

One way to tackle the problem of dust concentrations in pig housing is avoiding the peak concentrations that are caused by periods of agile movements of the pigs. However, to control the activity of the pigs there is still a lack of sensors to measure their activity level continuously in an on-line way. The purpose of this project was to develop a method for measuring the activity level of pigs in a pen in real time and identify an on-line behavior model to detect and predict occurrences of aggressiveness of pigs. For the experiments, 2 adjacent pig pens were used, each 6.9 by 2.6 meters in size and each populated with about 15 pigs. The equipment used for activity measurement consisted of an infrared-sensitive CCD camera that was mounted to the roof, 5 meters above the floor of the pen and connected to a PC with built in frame grabber. The camera was in a protective housing to shield it from dust and moisture. The lens was pointed downwards to get a top view of the pen. Images were captured with a resolution of 768 by 586 pixels and a 1 Hz frame rate. Software was developed to measure the activity level of animals, visible in the camera image, in real time and in practical conditions, i.e. a pigpen. Before the experiments, the two pens visible in the camera images were each divided in two zones, covering the left and right hand side, respectively. Every second, the algorithm logged the camera image and the activity index for each zone, defined as the fraction of the floor space in the pen that was covered or uncovered by pigs in the camera image during a one second period, i.e., the activity index is the fraction of floor space that is 'moving'. Finally the recorded video images were visually labeled in order to score animals' behavior and find a relation with the automatically measured activity index. From the analysis of the automatically measured group activity index compared to the manual labeling a relation exists between the activity index and the behavior types 'no activity', 'nuzzling' and feeding \ However, behavior related to aggressiveness of individual pigs ('fighting' and 'biting') can not be detected from instantaneous measurements of the group activity index.
Settore AGR/10 - Costruzioni Rurali e Territorio Agroforestale
ago-2008
American Society of Agricultural and Biological Engineers
CIGR
http://www.asabe.org/meetings/ILES2008/iles8abstracts.pdf
Book Part (author)
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/53521
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? ND
social impact