Image Recognition Study on Leaf Characteristics of Medicinal Plants Based on Otsu Dynamic Threshold Segmentation Algorithm

This experiment is designed to realize the recognition of different plants, and the main direction is to study based on the characteristics of plant leaves. First, the core technologies in the experiment, such as preprocessing, feature extraction, feature recognition and other related methods and algorithms, are introduced in detail, and relevant experiments are carried out. The experimental results show that the accuracy of recognition of some plants can reach about 96%, which is relatively high and can meet the development needs.