The objective of this study is to develop an algorithm for detecting oestrus in dairy cows from measurements of activity and duration of lying/standing periods. Each cows activity is measured by a sensor attached to the neck that returns an activity index for each hour. Duration of lying is measured by a sensor attached to the hind leg of the cow. Activity and lying/standing behaviour are modelled as a discrete event system, constructed using automata theory. In an attempt to estimate a biologically relevant lying balance, a lying balance indicator is constructed and is influencing transition probabilities in the stochastic automata. The cows lying-balance indicates how much the cow has been resting during the immediately past period, and the balance express to the automata, the tendency of the cow to continue resting or not. Automata for describing the two scenarios; normal and oestrus are designed and results of decision algorithms are presented for Oestrus detection. Detection based on the lying balance indicator and the two sets of measured information are demonstrated to increase the detection sensitivity to 100% for a set of 10 cows.
|Title of host publication||Precision livestock farming '09|
|Publication status||Published - 2009|
|Event||The Joint International Agricultural Conference - |
Duration: 1 Jan 2009 → …
|Conference||The Joint International Agricultural Conference|
|Period||01/01/2009 → …|
- animal monitoring