Prediction and determination of mildew grade in grain storage based on FOA-SVM algorithm

Grain mildew is a significant hazard that causes food loss and poses a serious threat to human health when severe.Therefore,effective prediction and determination of mildew grade is essential for the prevention and control of mildew and global food security.In the present study,a model for predicting and determining the mildew grade of rice was constructed using logistic regression,back propagation neural network and GS-SVM(a grid search-based support vector machine algorithm)based on laboratory culture data and actual data from a granary,respectively.The results show that the GS-SVM model has a better prediction effect,but the model cannot automatically adjust the parameters and is more subjective,and the accuracy may decrease when the data set changes.Therefore,this paper establishes a new model for a support vector machine based on a fruit fly optimization algorithm(FOA-SVM),which can achieve automatic parameter search and automatically adjust its parameters to find the best result when the data set changes,with a strong ability of self-adjustment of parameters.In addition,the FOA-SVM converges quickly and the model is stable.The results of this study provide a technical method for early identification of mildew grade during grain storage,which is beneficial for the prevention and control of rice mildew during grain storage.

Corn kernel classification from few training samples

This article presents an efficient approach to classify a set of corn kernels in contact,which may contain good,or defective kernels along with impurities.The proposed approach consists of two stages,the first one is a next-generation segmentation network,trained by using a set of synthesized images that is applied to divide the given image into a set of individual instances.An ad-hoc lightweight CNN architecture is then proposed to classify each instance into one of three categories(ie good,defective,and impurities).The segmentation network is trained using a strategy that avoids the time-consuming and human-error-prone task of manual data annotation.Regarding the classification stage,the proposed ad-hoc network is designed with only a few sets of layers to result in a lightweight architecture capable of being used in integrated solutions.Experimental results and comparisons with previous approaches showing both the improvement in accuracy and the reduction in time are provided.Finally,the segmentation and classification approach proposed can be easily adapted for use with other cereal types.