Chinese Herbal Classification Based on Image Segmentation and Deep Learning Methods

Abstract

AbstractThere are many kinds of Chinese herbal medicine and the same kind of Chinese herbal medicine will appear different forms due to the influence of growth environment and other factors. Furthermore, there are many different kinds of Chinese herbal medicine plants are very similar. Therefore it is a difficult task to classify them accurately. In recent years, deep learning has developed rapidly, and has achieved very good results in the field of object classification and recognition. So we decided to use GoogLeNet to classify the Chinese herbal medicine plants by their images. Because most of the images were taken in the natural state, so there will be very complex background inevitably, that is quite a difficult for classification. Here we developed the Chinese herbal medicine classification method based on image segmentation and deep learning methods. We first segment the images and then classify them. Experiment results show that we can get the accuracy of TOP-1 67.4% and TOP-5 92.4%. Through our experiments, we found that the combination of image segmentation and deep learning can achieve better classification results. This provides a good exploration for the field of image classification.KeywordsChinese herbal medicineImage segmentationClassification