AUTOMATED MONITORING TOOL TO DETECT THE SHORT-TERM FEEDING BEHAVIOURS OF BROILER CHICKENS BY SOUND ANALYSIS

摘要

This paper describes a novel monitoring system to accurately detect the short-term feeding behaviours (meal size, meal duration, meals per day and feeding rate) of broiler chickens at group level by a real-time sound processing technology. In this research, the pecking sounds of 10 male, 39 days old, broiler chickens were recorded by a microphone that was attached to the feeder. Simultaneously, the appearances of broilers around feeder were recorded by a camera that was positioned at the top of the feeding pen. At the same time, a weighing system was used to automatically record feed uptake of broilers as a reference method. Moreover, an existing algorithm was used to detect the pecking sounds of 10 broiler chickens while the birds were all eating together. The feeding behaviours of broiler chickens were obtained by pecking sound analysis. The results of the algorithm were compared to reference feeding behaviours through weighing system measurements and video observations. The relationship betweenfeeding behaviours obtained by algorithm and feeding behaviours recorded by a weighing scale and video camera was investigated and a strong positive correlation was found between these methods. Furthermore, the linear regression tests which were performed which were performed resulted in the following coefficients of determination R2 = 0.965 for meal size, 0.938 for meal duration, 0.896 for the number of meals per day and 0.888 for feeding rate. Additionally, the estimated accuracy of the method showedthat, 90% of meal size, 95% of meal duration, 94% of the number of meals per day and 89% of feeding rate were correctly monitored using sound analysis. The results suggest that this automated continous measurement system has the potential to be used as atool to monitor the short-term feeding behaviours of broiler chickens at group level. The most important advantage of this system is that the measurements can be made continuously throughout the life span of a flock, in a fully automated, completely non-invasive and non-intrusive way. It is proposed to test this system in commercial conditions due to its low costs and in order to evaluate its applicability under these conditions. Future researchs should be focused on sound technology to assess the health and welfare of broilers by automatically and continouslly monitoring the feeding behaviours.