Design of plant intelligent recognition system based on transfer learning

The identification of plant species relies heavily on flowers of plants, which are crucial for the process. Traditional manual methods for identifying plants are not only labor-intensive and inefficient, but also struggle to ensure accurate data. With the advancement of artificial intelligence technology, mobile software can now quickly identify plants, leading to widespread use in agriculture and ecology. This study utilizes transfer learning to significantly enhance the accuracy and efficiency of plant recognition. It introduces the current research status of plant recognition, conducts a comparative analysis of mainstream convolutional neural network models, and uses VGG16 and ResNet-50 pre-trained models for feature extraction and classification. Experimental results using five different flower datasets demonstrate high accuracy for both methods. Additionally, a visual interface based on the model is created for efficient plant identification with low computational cost.