一种基于卷积神经网络的烟叶等级识别方法  

A method of tobacco leaf grade recognition based on convolutional neural network

作　　者：焦方圆 申金媛[1] 郝同盟 JIAO Fang-yuan;SHEN Jin-yuan;HAO Tong-meng(Zhengzhou University,Zhengzhou,Henan 450001,China;North China University of Water Conservancy and Hydropower,Zhengzhou,Henan 450045,China)

机构地区：[1]郑州大学,河南郑州450001 [2]华北水利水电大学,河南郑州450045

出　　处：《食品与机械》2022年第2期222-227,共6页Food and Machinery

基　　金：国家自然科学基金(编号:69587005)。

摘　　要：目的:解决烟叶分级准确率不高的问题。方法:提出一种改进的基于卷积神经网络的烟叶分级模型,根据VGG16网络结构,以自定义的方式搭建网络模型;将空洞卷积代替原有的传统卷积,增加图像感受野的同时避免了图像特征的损失,并将激活函数改为Leaky;elu,修正数据的分布,解决ReLU函数的硬饱和问题;用41种等级的烟叶图片加以测试。结果:试验改进算法分级准确率达95.89%,与传统SVM算法相比提高了10.46%,与经典VGG16算法相比提高了7.87%,损失率最终收敛于0.13。结论:与原始模型和传统特征提取的方式相比,试验算法在烟叶分级准确率性能上有所提高。Objective: To solve the problem of low accuracy of tobacco grading. Methods: An improved tobacco leaf grading model based on convolutional neural network was proposed. According to the VGG16 network structure, the network model was built in a custom way. The traditional convolution was replaced by the hole convolution, which increased the image receptive field while avoiding. The loss of image features was changed, and the activation function was changed to Leaky;elu. The data distribution was corrected, and the hard saturation problem of the ReLU function was solved. 41 levels of tobacco leaf pictures were used for testing. Results: The grading accuracy rate of the test algorithm was 95.89%, which was 10.46% higher than the traditional SVM algorithm, and 7.87% higher than the classic VGG16 algorithm. The loss rate finally converged to 0.13. Conclusion: Compared with the original model and traditional feature extraction methods, this algorithm has improved the accuracy of tobacco leaf classification.

关 键 词：烟叶分级 深度学习 激活函数 空洞卷积 

分 类 号：TS42[农业科学—烟草工业] TP183[自动化与计算机技术—控制理论与控制工程] TP391.41[自动化与计算机技术—控制科学与工程]