基于轻量化YOLO模型的植物叶片识别

Plant leaf recognition based on lightweight YOLO model

为了提升植物叶片的检测能力,减少繁琐的人工成本,本文构建了包含10类常见植物叶片的数据集。应用基于pytorch框架的YOLOv5网络模型对叶片数据集进行训练,并通过训练模型对叶片测试图像进行检测,在识别速度和精度上取得了较好的效果。实验结果表明,本文检测方法对叶片识别的平均精度为93%,识别速度较快,有效解决了传统植物识别方法中分类器耗时长、准确率低等问题。

In order to improve the detection capability of plant leaves and reduce the tedious labour cost, a dataset containing 10 types of common plant leaves was constructed in this paper. A YOLOv5 network model based on the pytorch framework was applied to train the leaf dataset, and the leaf test images were detected by the training model, which achieved better results in terms of recognition speed and accuracy. The experimental results show that the average accuracy of the detection method in this paper for leaf recognition is 93% and the recognition speed is fast, which effectively solves the problems of time-consuming classifier and low accuracy rate in traditional plant recognition methods.