Design And Development Of Pistachio Foreign Body Automatic Sorting System

Impurity detection is an important link in the food processing process.Taking pistachios as an example,it is a big problem to accurately detect and remove foreign matter mixed in pistachios before packaging.In order to accurately identify and remove foreign objects mixed in pistachios,this paper designs an automatic pistachio sorting system based on deep learning.The specific research contents and results are as follows:This paper constructs a data set based on pistachio and related foreign objects,and uses data enhancement to expand the sample data set;researches and analyzes the current mainstream deep learning target recognition algorithm,and chooses YOLOv5 as the basis of the pistachio foreign object sorting and detection model in this paper web framework.Construction of pistachio foreign body detection model based on YOLOv5 network model.A small target detection layer is added to the backbone to improve the model’s ability to detect small foreign objects and broken shells;a Tans-E module is proposed to replace some convolution blocks in the original network,which can better capture the global information of the image;Adding the CBAM module improves the sensitivity of the model to channel features,making it easier to find prediction areas in dense objects;modifying the IOU loss function to EIOU makes the convergence effect better;finally introducing the Ghost Net module to make the model lighter and improve the model detection rate.The test results show that the m AP of the improved YOLOv5 model reaches 88.9%,which is 6.7%higher than the average accuracy of the original YOLOv5 network.The recognition effect of small impurities is significantly better than the original YOLOv5 network,and the detection rate is only reduced by 3.9%.Based on NX secondary development technology,the establishment of pistachio sorting device model and the design of parametric design system were realized.Created NX customized menu bar by using Menu Script tool,completed the design of parametric design system dialog box based on MFC,adopted top-down parametric design method to transfer model parameters,and used C++ language to write parametric modeling program,realized When the design parameters are changed,the pistachio foreign matter sorting mechanism model is changed accordingly.The mechanical mechanism of the sorting equipment was designed and manufactured in detail,and the related control system was designed to control the sorting device and realize the pistachio foreign matter sorting function.And the same sample as the data set was used for the test,and the recognition effect of the model on the data set samples was verified through the sorting platform.The recognition test results showed that the pistachio foreign body detection model based on the improved YOLOv5 established in this paper has good accuracy and average precision.Finally,pistachio materials were used to separate foreign bodies,with an average success rate of 97.5%.