Classification of Hyperspectral Milk Varieties Based on SSA-BP Neural Network

Abstract

Based on hyperspectral technology, a BP neural network classification model based on sparrow algorithm was proposed to solve the shortage of fast classification technology of milk. The experimental samples were five kinds of milk with different nutrient contents, and the wavelength data within 400-1000nm were collected. The competitive adaptive weighted sampling algorithm and continuous projection algorithm were used to extract the characteristic wavelength from the pre-processed spectral data, and the sparrow algorithm was used to optimize the BP neural network to find the optimal weight threshold training model. The classification results were compared with the support vector machine model and BP neural network model. The results showed that the training set accuracy of CARS-SSA-BP model was 100%, and the test set accuracy was 98.14%. The classification effect of CARS-SSA-BP model was better than that of SVM model and BP neural network model. Sparrow algorithm can effectively optimize the weight and threshold value of BP neural network, and it is feasible for the identification of milk varieties.