Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Using Decision Tree for Classifying Farrowing Piglets Stress Conditions using Vocalization

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  2012 IX International Livestock Environment Symposium (ILES IX)  ILES12-0895.(doi:10.13031/2013.41534)
Authors:   Alexandra F. S Cordeiro, Irenilza A Nääs, Stanley R.M Oliveira, Andréia C.M Almeida, Diego P Neves
Keywords:   Farrowing sows, animal welfare, signal analysis

The objective of this study was to classify different situations of stress from the vocalization of pigs in the phase of motherhood. Vocalization of 40 pigs was recorded in different stress situations in the farrowing phase (without stress, heat stress, pain, restrained from nursing) . To capture the acoustic signals unidirectional microphone was used at approximately 15 cm from the mouth of the piglets. This microphone was connected to a digital recorder, and the signals were digitized at a frequency of 44,100 Hz. Sound signals were edited and analyzed using the Praat software. Each collected signal was divided into three samples, and from each sample, it was extracted the maximum and minimum amplitudes, the frequency of Pith; and the frequency of formants 1, 2, 3 and 4, was obtained using that attributes within the range of amplitude (the difference of the amplitudes maximum and minimum), and duration of the signal. The algorithm J48 of Weka software was used for determining the stress condition. It was possible to classify four levels of stress: no-stress, pain, cold and hunger from the vocalization of pigs to the farrowing phase, with 77.95% accuracy.

(Download PDF)    (Export to EndNotes)