摘要
吐丝期是玉米生长期的关键节点,对其识别至关重要。基于此,在Mobilenetv2模型的基础上进行改进,并在玉米数据集上进行迁移学习,对其网络参数进行训练、微调,得到精准的玉米吐丝期识别模型。实验证明对Mobilenetv2模型进行迁移学习在玉米吐丝期识别上具有良好的效果,比传统方法有显著提高。
The silk stage is the key node of maize growth period,and it is very important to identify it.In this paper,the Mobilenetv2 model is improved,and the transfer learning is carried out on the corn data set,and the network parameters are trained and fine-tuned to obtain an accurate corn spinning-stage recognition model.The experimental results show that the transfer learning of Mobilenetv2 model has a good effect on the identification of maize spinning-stage,which is significantly higher than the traditional method.
作者
李涛
杨凡稳
张震
贾红宾
Li Tao;Yang Fanwen;Zhang Zhen;Jia Hongbin(College of Information Engineering,Xinxiang Vocational&Technical College,Xinxiang Henan 453006,China)
出处
《山西电子技术》
2025年第5期11-13,共3页
Shanxi Electronic Technology
关键词
玉米
吐丝期
深度学习
迁移学习
maize
silking stage
deep learning
transfer learning
