Application of Wavelet Transform and Neural Network in Near Infrared Spectroscopy Analysis in Pork

In the present study, an innovative method is proposed, employing both wavelet transform and neural network, to analyze the near-infrared spectrum data in pork. The method entails using db3 wavelet at 5 levels decomposition to process original spectral data, using the transformed data as the input matrix, and creating the model through neural network. The PLS model and ANN model are established by using the original spectral data respectively, and the neural network model is established by using the wavelet feature data. In contrast to the third models can be seen wavelet neural network modeling is the best, and the root-mean-square error of prediction (RMSEP) of moisture, fat and protein reached 0.1808%, 0.1087%, 0.2953%, and correlation coefficient (R) of the moisture, fat, protein were 0.9863, 0.9945, 0.9593. The results show that the wavelet neural network method provides a new method for the detection of moisture, fat, protein and high precision in pork.