Tea categories classification using morphological characteristic andsupport vector machines

摘要

Tea categories classification is an importance task for quality inspection. And traditional way for dong this by human istime-consuming, requirement of too much manual labor. This study proposed a method for discriminating green teacategories based on multi-spectral images technique. Four tea categories were selected for this study, and total of 243multi-spectral images were collected using a common-aperture multi-spectral charged coupled device camera with threechannels (550, 660 and 800 nm). A compound image which has the clearest outline of samples was process bycombination of the three monochrome images (550, 660 and 800 nm). After image preprocessing, 18 morphometryparameters were obtained for each samples. The 18 parameters used including area, perimeter, centroid and eccentricityet al. To better understanding these parameters, principal component analysis was conducted on them, and score plot ofthe first three independent components was obtained. The first three components accounted for 99.02% of the variationof original 18 parameters. It can be found that the four tea categories were distributed in dense clusters respectively inscore plot. But the boundaries among them were not clear, so a further discrimination must be developed. Threealgorithms including support vector machines, artificial neural network and linear discriminant analysis were adopted fordeveloped classification models based on the optimized 9 features. Wonderful result was obtained by support vectormachines model with accuracy of 93.75% for prediction unknown samples in testing set. It can be concluded that it is aneffective method to classification tea categories based on computer vision, and support vector machines is veryspecialized for development of classification model.