| China is a large poultry consumption country,poultry wing segmentation is an important process in poultry slaughtering and processing.However,due to the large difference of poultry body,the traditional manual operation is still adopted in most small and medium-sized enterprises.Therefore,it is of great significance to develop an automatic poultry wing segmentation system which can adapt to different types of poultry to speed up the modernization of poultry slaughtering and processing.In this paper,Sanhuang chicken as the research object,comprehensive analysis of domestic and foreign research status of bird wing segmentation,design a set of automatic segmentation system based on machine vision bird wing.The system is mainly composed of mechanical device,control system and vision system.In this paper,the design and construction of the three parts of avian wing segmentation are carried out.The main research contents are as follows:(1)bird wing segmentation mechanism.The mechanical devices designed in this study mainly include poultry carcass conveying device,adjustable bird body clamping component,bird wing segment mechanical arm component and arm end cutting device.The position of bird wing cutting on the poultry transport line was obtained through the visual system,and the mechanical arm carrying the milling cutter ran synchronously with the poultry carcass to realize the online bird wing segmentation operation.The system can segment the wings of different kinds and different body types of poultry.(2)Visual system of avian wing segmentation.The visual system is composed of industrial camera,lens,light source and visual controller.OpenCV4.5.5 machine vision library is used to complete the image processing of bird wing segmentation on VS2015.After the original image is gray-scale,Gaussian filter denoising and image enhancement preprocessing,adaptive threshold algorithm,canny edge detection algorithm and K-means clustering algorithm are used to segment the preprocessed image respectively.The results show that the adaptive threshold algorithm has the best segmentation effect of bird wings,and the segmentation accuracy reaches 91.42%.The running time was reduced to 8.61 s.The regional growth algorithm was used to extract the bird wing part,obtain the coordinates of the segmentation points,and transmit the coordinates information of the segmentation points to PLC.(3)avian wing segmentation control system.Mitsubishi PLC as the main body to control the automation system,according to the requirements of the manual and automatic operation of two control modes,according to the avian wing segmentation process requirements to determine the robotic arm end cutting device running path,and the use of ladder diagram in the form of PLC programming,At the same time,MX Component software is used to complete the communication between PC and VS2015 software,and appropriate human-computer interaction interface is designed to facilitate human-computer interaction.(4)bird wing segmentation prototype test.The rationality of the mechanical structure,the feasibility and accuracy of the vision system and the stability of the control system are verified.The results showed that the average residual rates of left wing and right wing were 9.15% and 10.42%,respectively,and the optimal residual rates were as low as 8.27% and 8.21%,respectively.The overall average residual rate of bird wings was 9.79%,and the optimal residual rate of monomer was as low as8.63%,which met the requirements of this paper.In this study,the automatic poultry wing segmentation operation was closely combined with the research on bird wing segmentation method based on machine vision,which basically achieved the segmentation effect required by the test,and provided technical support for the fully automated operation of poultry slaughter production line. |
PDF Full Text Request