Abstract:In the practice of forensic science, it is often necessary to accurately determine the identity of the test material and the sample document by analyzing the composition of the ink in the document.The hyperspectral imaging technology combined with machine learning was used to identify the types of inkjet printing inks. A method for classifying inkjet printing inks is presented by using hyperspectral imaging and spectrophotometry combined with chemometry.The hyperspectral data and chroma values of 14 sets of four color inkjet printing inks of different brands and models were collected.After calculating the average chromaticity value, PCA dimensionality reduction and K-Means cluster analysis were performed to classify the samples.Then three classification models, LightGBM model, XGBoost model and SVM model, were used to determine the test set and training set at a ratio of 1:4, and each class of samples in the cluster analysis results were identified one by one.The results show that the classification accuracy of both LightGBM and XGBoost can reach more than 95%, and the classification accuracy of SVM is 100%.The proposed method can distinguish different brands and models of inkjet printing inks without damage, accurately and quickly.