Computational method for segmentation and classification of ingestive sounds in sheep
Computers and Electronics in Agriculture
Automatic recognition of ingestive sounds of cattle based on hidden Markov models
Computers and Electronics in Agriculture
Computers and Electronics in Agriculture
| Hi-index | 0.00 |
Estimating forage intake by free-grazing livestock is difficult and expensive. Previous approaches include behavioral observation, ratio techniques using indigestible markers, mechanical recording of ingestive jaw motion, and acoustic recording of ingestive behaviors. Acoustic recording shows great potential but has been limited by the difficulty and time required to manually identify and classify ingestive events. We present an acoustic recording and analysis system that automatically detects, classifies, and quantifies ingestive events in free-grazing beef cattle. The system utilizes a wide-frequency acoustic microphone close to the animal's mouth, mathematical signal analysis to detect and measure ingestive events, and streaming data analysis capable of handling an unlimited amount of data. Analysis parameters can be reconfigured for different animals, forages and other changing conditions. The system measures the acoustic parameters of ingestive events, such as duration, amplitude, spectrum and energy, which can support further event classification and become the inputs to a forage intake model. We validated our detection and classification technique against the results of trained human observers based on field studies with grazing steer. The software detected 95% of manually identified bites in an event-by-event comparison. Field observations and sound attenuation analysis indicate that sounds from adjacent livestock and ambient pastoral environments have an insignificant effect upon the integrity of the recorded acoustic data set. We conclude that wideband acoustic analysis allows us to identify ingestive events accurately and automatically over extended periods of time.