Research on Accurate Flower Species Recognition Technology Based on ResNet Algorithm

This paper aims to solve the problem of accurate identification of flower varieties. We come up with a convolutional neural network model based on deep learning framework, which is ResNet18 neural network model, from the multiple aspects, such as network degradation, model training, and recognition accuracy. This paper conducts simulation experiments on the identification of flower varieties by using the ResNet18 network model, achieving the reduction of the impact of network degradation, the improvement of the training efficiency of the ResNet18 model, and a relatively high accuracy rate in identifying flower varieties. This paper yield positive result by analyzing the experiment data. Therefore, the simulation experiment shows that the ResNet18 network model has achieved success, and the ResNet18 model provides a solution for solving the problem of flower variety identification.