Frequency analysis of vocalisations to monitor broiler chicken production performance in real-life farm conditions
Contributo in Atti di convegno
Data di Pubblicazione:
2017
Citazione:
Frequency analysis of vocalisations to monitor broiler chicken production performance in real-life farm conditions / L. Carpentier, T. Norton, D. Berckmans, I. Fontana, M. Guarino, E. Vranken, D. Berckmans - In: Precision Livestock Farming 2017[s.l] : European Conference on Precision Livestock Farming, 2017 Sep. - pp. 138-146 (( Intervento presentato al 8. convegno European Conference on Precision Livestock Farming tenutosi a Nantes nel 2017.
Abstract:
In today's broiler houses, birds are reared in large groups, with up to 50,000 often together in one house. Supervising all these animals is difficult or even impossible for the farmer. Sensors like microphones can help the farmer by continuously monitoring the productivity of the animals and raising early alarms in the case abnormalities in the house. As regards linking sound to production performance, recent studies indicate that there is an inverse relation between the peak frequency of animal sounds and their weight. This study aims to monitor the frequencies of the vocalisations in a broiler house and compare these to the weight evolution of the chickens. The resulting curves compare different production cycles with the aim to provide the farmer with a tool to compare the performance of the current and previous production cycles. Pre-processing was performed by filtering the data in order to remove the machinery noise. After this the median frequency was extracted from the data. It was found that between day 3 and 25, there was a clear difference in frequency values between a production cycle with heavier birds (lower frequency values) and a production cycle with lighter birds (higher frequency values). Based on these results, there is an indication that the method developed has potential to differentiate between light and heavy flocks in a commercial broiler house. Comparing the different batches using the developed approach can help the farmer to distinguish a good and a bad production cycle early and assist the farmer in adapting his management practice.
Tipologia IRIS:
03 - Contributo in volume
Keywords:
Bioacoustics; Precision livestock farming; Signal processing; Spectral analysis
Elenco autori:
L. Carpentier, T. Norton, D. Berckmans, I. Fontana, M. Guarino, E. Vranken, D. Berckmans
Link alla scheda completa:
Titolo del libro:
Precision Livestock Farming 2017