基于深度学习的甘蔗茎秆切面检测程序设计与开发

Design and Development of Sugarcane Stem Section Detection Program Based on Deep Learning

为了实现对甘蔗茎秆切面的自动检测,笔者设计并开发了一款基于深度学习的甘蔗茎秆切面检测小程序。首先,将收集的数据集作为训练数据集;其次,使用UNet++网络结构训练模型;最后,利用训练好的模型进行微信小程序的开发,将训练好的模型嵌入微信小程序后端,设计并开发甘蔗茎秆切面检测小程序。结果表明,该程序具有页面简洁明了、方便用户操作等特点,其可视化分析大大缩短了用户对数据的分析周期,可为农业植物生产领域的研究人员提供甘蔗茎秆切面的相关检测数据。

In order to realize the automatic detection of sugarcane stem section,the author designs and develops a small program for sugarcane stem section detection based on deep learning.Firstly,the collected data set is used as the training data set.Then train the model using UNet++network structure.Finally,the well-trained model is used to develop the WeChat mini-program.The well-trained model is embedded in the back end of the WeChat mini-program,and the sugarcane stalk section detection mini-program is designed and developed.The results show that the program has the characteristics of simple and clear pages and convenient operation for users.Its visual analysis greatly shortens the analysis cycle of users for data,and can provide researchers in the field of agricultural plant production with relevant detection data of sugarcane stalk sections.