Establishment Of Chicken Physical Health Model Based On Machine Vision And Pressure Reaction

People's dietary structure and consumption have been largely changed with the increase of economics. More attention of people are inlined to their health, as well as the effect of diet on health. The occurance of many disease has been proved to be related with diet and dietary structure, like some particularly zoonoses, including mad cow disease, avian-influenza, and so on. People have more concerns about chicken's physical healthy and quality with large scale farming. In addition, application of genetic engineering technology in chicken improves the production efficiency and performance, while it also makes chicken more susceptible to disease. The physical health of poultry is related to the morbidity and mortality of chicken, also directly affects the benefit of the breeding enterprise.At present, effective methods for detecting chicken physical health have been not reported. In our study, the “Meihuang 2” chicken were chosen as my research object, and machine vision and pressure reaction were used to comparephysical health of chicken reared under five differentconditions. The main research contents and conclusions are as follows,(1) Establishment of health model for chicken physical based on the machine vision:The difference of chickens grow performance were analysed when five different condition of rearing were used. And the video image in drop test of chicken using machine vision was acquired and analysed to get the number of wing waving in the drop processing. Finally, the health model of chicken's physical was established using the number of wing waving, day age and weight of chicken. The results showed that chicken's weight negatively correlated with the starvation time, and chicken's weight is more sensitive response to starvation time with chicken's day ageraising. Meanwhile, to encode the video image of chicken drop processing by using machine vision technology, automatic extraction the wings of chicken drop processing, the number of wing waving changed with day age and weight. Combined the number of wing with chicken's day age and weight to establish physical health Logistic model of chicken, the model can be used to classify two kinds of chicken, “normalrearing” and “hunger feeding”with classification accuracy of 84%.(2) Establishment of physical health model for chicken based on the pressure reaction of chicken: This study proposed a method based on the mechanical properties or pressure reaction to evaluate health index of chicken body. Firstly, the pressure platform of universal testing machine slow down to put pressure on the chicken, and the pressure curve and effective parameters were recorded in the process of chickens yield. Secondly, maximum resistance of chickens yield, chicken day age and weight were used to establish health model of chicken's physical. The results showed that the correlation between the body weight and yield force of chicken with the chicken day age had obvious linear relationship, and that chicken's yield force negatively correlated with the starvation time. Meanwhile, chicken's yield force is more sensitive response to starvation time with chicken's day age raising. Combined chicken's maximum resistance withchicken's day age and weight, establish physical health model of chicken by using the Logistic analysis method, the “normal rearing” and “hunger feeding” classification accuracy of 83.2%.