Peter-Christian Schön

Linear prediction coding analysis and self-organizing feature map as tools to classify stress calls of domestic pigs (Sus scrofa)

It is assumed that calls may give information about the inner (emotional) state of an animal. Hence, in the last years sound analysis has become an increasingly important tool for the interpretation of the behavior, the health condition, and the well-being of animals. A procedure was developed that allows the characterization, classification, and visualization of the cluster structures of stress calls of domestic pigs (Sus scrofa). Based on the acoustic model of the sound production the extraction of features from calls was performed with linear prediction coding (LPC). A vector-based self-organizing neuronal network was trained with the determined LPC coefficients, resulting in a feature map. The cluster structure of the calls was then visualized with a unified matrix and the neurons were labeled for their input origin. The basic applicability of the procedure was tested by using two examples which were of special interest for a possible evaluation of the normal farming practice. The procedure worked well both in discriminating individual piglets by their scream characteristics and in classifying pig stress calls vs other calls and noise occurring under normal farming conditions.

A SOUND ANALYSIS SYSTEM BASED ON LABVIEW® APPLIED TO THE ANALYSIS OF SUCKLING GRUNTS OF DOMESTIC PIGS SUS SCROFA

ABSTRACT The paper gives an overview of the capacities of the software solutions for a sound analysis system based on the program development environment LabVIEW and shows some applications of the system by using suckling grunts of domestic pigs Sus scrofa as an example. LabVIEW represents a versatile tool for the development of sound analysis software including sound processing and sound statistics. Particularly, the increasing need for flexible numerical processing of different acoustic parameters can be considered. The graphical object oriented language allows programming without special experience because it uses terminology, icons and theories familiar to scientists and engineers. LabVIEW is available for Macintosh, Sun, HP-UX, Windows 95, Windows NT, Windows 3.1., and can also be used as a teaching tool to gain technical knowledge of signal analysis and signal processing. Paralleled by behavioural observations, the bioacoustical studies of animals at various levels, including the incorporation of te...

Objectively measuring behaviour traits in an automated restraint-test for ungulates: towards making temperament measurable

The personality of an animal is described by traits that cause consistent actions and reactions to environmental stimuli. An important part of personality is the reaction to unpleasant or uncontrollable situations. Methods described in the literature to measure personality in animals are often based on measuring or rating escape behaviour in these situations. In the methods described, human handlers are frequently part of the experiment or the animals’ personalities are scored by humans. Thus, these methods are at least partly subjective. In the current study, an appliance to measure objectively the escape behaviour of ungulates and their reluctance during an uncontrollable situation (restraint) with a rather simple and comprehensible methodology is presented using a force transducer with adequate peripheral equipment. While the animals were restrained, a tractive force-time diagram describing escape behaviour was recorded and later analysed with software developed specifically. To evaluate this newly developed technical method, 24 three-month-old calves were restrained by being tethered for 30 min on a halter that was connected to the force transducer. From the tractive force-time diagram, tractive force, maximal tractive force and the number of pulls that the calves performed during 5-min intervals were calculated. The multivariate results were analysed with a k -means-algorithm (function ‘kcca’) and a hierarchical clustering (function ‘hclust’) included in R version 2.12.1. Both analyses revealed two clearly separated clusters including the same individuals in each analysis. The animals of cluster 1 showed a continuously higher reaction level than those of cluster 2 with a strong reaction in the beginning, a short decrease before increasing during the middle of the experiment and a final decrease at the end of the test. The animals of cluster 2 had a lower and quite steady reaction level throughout the experiment, although even here a slight increase during the middle of the experiment could be detected before a final decrease towards the end of the test was shown. There was no significant difference in weight between the two clusters. The results showed that this newly developed method was able to detect differences in the animals’ escape behaviour patterns and reluctance with the measured parameters.