Design And Research Of Auxiliary Device For Sheep Face Image Acquisition

Identification of sheep by sheep face image is a new technology of individual identification of sheep,and is the basis of accurate animal rearing and management.The technology has the advantages of non-contact and non-invasive,which can avoid contact between humans and animals,make sheep produce less stress response,and avoid zoonosis.Sheep face image acquisition is basic,but also difficult.Therefore,this thesis designed and studied the auxiliary device for sheep face image acquisition,and the work and achievements are as follows:(1)Overall part: I designed the frame of the whole device,and finally determined that the overall frame of the device is composed of 8 square tubes and 2 bottom plates.The sheep part of the transmission using pneumatic transmission.Determine the frame size of the whole device,refer to relevant literature and books,understand the size and weight of each breed of sheep,so as to determine the size of the whole device components.(2)The mechanical part of the device is designed and studied: the purpose of the design and study of the device is to fix the sheep by holding the sheep,so the clamping part is made of HQ60 steel plate in line with the sheep belly curve,and the movement of the clamping part and the sheep into the front and back doors of the device is driven by the cylinder.(3)The control part of the device is designed and studied.Pneumatic part according to the known most vigorous selection of sheep face recognition auxiliary device cylinder and air pump.Omron CP1H-X40DT1-D PLC is selected for the electronic control part of the whole device.(4)Prototype production and test: the clamping device of the auxiliary device is simulated and analyzed.After meeting the requirements,we processed the equipment.After processing,we have connected the control section.Finally,the device was tested in a shed where the sheep were raised and the facial image of the sheep was collected successfully.A data set of 2260 images was generated,which could be used to train and test the model.