Integrating radio frequency identification into the PiGUI system to recognise sampling bias and detect feeding behaviour

Tscharke, M. and Banhazi, T. M. (2013) Integrating radio frequency identification into the PiGUI system to recognise sampling bias and detect feeding behaviour. Australian Journal of Multi-disciplinary Engineering, 10 (1). pp. 94-107. ISSN 1448-8388

Abstract

A machine vision system was developed to determine the live weight and growth rate of groups of pigs. During development, it was important to determine whether the sampling method had potential to cause bias and subsequent error in the daily weight estimates calculated by the system. Sampling bias can occur toward certain pigs because of their appearance or the frequency and duration in which they reside in the pen-region beneath the camera. To determine whether these forms of bias could occur, radio frequency identification (RFID) was integrated into the system to monitor the attendance of individual pigs in the pen-region observed by the camera. Test results indicated that both forms of bias had occurred as a result of the system’s filter settings and the installation position. As the system observed a single feeder space, the opportunity arose to analyse the data further and determine whether the feeding behaviour of individual animals could be recovered from their attendance at the feeder. Preliminary findings indicate that the attendance recorded by the RFID system at the feeder is related to weight gain and that attendance might be useful in detecting feeder demand and out of feed events. In addition, it is believed that the RFID-recorded interactions between pigs at the feeder may provide a novel way of automatically recording competitive behaviour between individual animals in a group. Continuously identifying individual pigs at the feeder helps to fine tune the vision systems parameters to overcome bias related issues concerning layout and sampling. Additional information can be gained by the RFID system which prompts further investigation.


Statistics for USQ ePrint 36775
Statistics for this ePrint Item
Item Type: Article (Commonwealth Reporting Category C)
Refereed: Yes
Item Status: Live Archive
Additional Information: Files associated with this item cannot be displayed due to copyright restrictions.
Faculty/School / Institute/Centre: Current - Faculty of Health, Engineering and Sciences - School of Civil Engineering and Surveying (1 July 2013 -)
Faculty/School / Institute/Centre: Current - Faculty of Health, Engineering and Sciences - School of Civil Engineering and Surveying (1 July 2013 -)
Date Deposited: 09 Aug 2019 04:50
Last Modified: 09 Aug 2019 05:53
Uncontrolled Keywords: machine vision, weight estimation, pigs, precision livestock farming, RFID, animal behaviour
Fields of Research : 07 Agricultural and Veterinary Sciences > 0702 Animal Production > 070203 Animal Management
Socio-Economic Objective: E Expanding Knowledge > 97 Expanding Knowledge > 970107 Expanding Knowledge in the Agricultural and Veterinary Sciences
Identification Number or DOI: 10.7158/14488388.2013.11464868
URI: http://eprints.usq.edu.au/id/eprint/36775

Actions (login required)

View Item Archive Repository Staff Only