BUILDING A VIETNAMESE MEDICINAL HERB SEARCH SYSTEM

Abstract

People's living conditions are steadily improving in today's society. Simultaneously, various difficulties in the diet have emerged, leading to an increase in the occurrence of diseases. Herbal medicine from Vietnam has long been used to treat a variety of ailments. However, it is an issue in terms of collecting and classifying Vietnamese herbal remedies. There are many different kinds of Vietnamese herbal medicine plants, and some of them seem remarkably similar. Even a taxonomist has trouble distinguishing between all herbal medicines, let alone a novice. So, The use of technology to recognize medicinal plants in the natural environment should be studied in light of the preceding context. I photographed 30 different species of Vietnamese medicinal plant leaves for this thesis. A total of 3,000 images were collected, with 100 images for each type. The data was analyzed using the deep learning method, and the accuracy of the results of the model is 92%.