2DCOS combined with CNN and blockchain to trace the species of boletes

Abstract

Bolete mushrooms are well received by consumers for their rich nutrition and high medicinal value. However, the nutritional value and medicinal value of different species of bolete mushrooms are significantly different. Therefore, it is necessary to identify and trace the species of bolete mushrooms. In this study, Support Vector Machine (SVM) model and four deep learning models with different data sets were established to identify the species of boletes. By comparison, the accuracy of the train set, test set and external verification can reach 100% about the synchronous two-dimensional correlation spectroscopy (2DCOS) model, and the loss value of this model is 0.0257 which is close to zero. Therefore, the synchronous 2DCOS model has the best accuracy and generalization ability. Then, the results of species identification were uploaded to the blockchain platform that we build. Users can query and display the information after identity authentication, so as to realize the traceability of bolete mushrooms. The results show that our method is feasible. The traceability technology based on deep learning and blockchain has been used in the field of microbiology in this research, and it can be extended to other fields.