Real time computer stress monitoring of piglets using vocalization analysis

The analysis of animal vocalization allows the interpretation of stress pattern in a non-invasive way, and it may be used to assess welfare as well as health status and social adaptation. This paper presents the development of software to monitor and analyze distinct sounds emitted by piglets correlating the noise response with levels of stress for assessing welfare. The software was developed using Delphi5. Through the component UwaveIn the sound was recorded with an electronic microphone and set to a PC hardware that registered the signal of amplitude noise intensity, expressed in decibel (dB), converting them into bytes. Fourier Transform was used to describe the sound spectrum and transform it into frequency and intensity. Validation of the software was done in a commercial swine farm using piglets from sows’ nursery. Five examples of vocalization were studied related to distinct levels of stress (fear and alertness) due to regular management procedures. The real time monitoring allowed the detection of piglet stress exposure through the surface point polynomial adjustment. The acoustic signal recorded in the software could be partitioned allowing an evaluation of the sonograph through time.