Development of a leaf recognition system based on convolutional neural networks

Plant leaves are crucial for plant classification, and their automatic recognition has broad applications in botany and agriculture. However, traditional leaf classification methods are often inefficient and inaccurate. This study proposes a leaf recognition system based on Convolutional Neural Networks (CNNs). Utilizing MobileNetV2 as the base model, the system accurately classifies different types of leaves through extensive training on leaf image data. Specifically, the system preprocesses the dataset and leverages CNN's powerful feature extraction capabilities to identify patterns like texture and shape. Validation on a test set achieved an average recognition accuracy of 96%, indicating high reliability. Additionally, the system features a visualizati on interface for users to access detailed recognition results. This research provides an effective solution for the rapid and accurate identification of plant leaves, with significant theoretical and practical implications.