[1]陆煜,俞经虎,朱行飞,等.基于卷积神经网络的轻量级水稻叶片病害识别模型[J].江苏农业学报,2024,(02):312-319.[doi:doi:10.3969/j.issn.1000-4440.2024.02.013]
　LU Yu,YU Jing-hu,ZHU Xing-fei,et al.A lightweight rice leaf disease recognition model based on convolutional neural network[J].,2024,(02):312-319.[doi:doi:10.3969/j.issn.1000-4440.2024.02.013]
点击复制

基于卷积神经网络的轻量级水稻叶片病害识别模型()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2024年02期
页码:
312-319
栏目:
农业信息工程
出版日期:
2024-02-25

文章信息/Info

Title:
A lightweight rice leaf disease recognition model based on convolutional neural network
作者:
陆煜1; 2; 俞经虎1; 2; 朱行飞1; 2; 张不凡1; 2
(1.江南大学机械工程学院,江苏无锡214122；2.江苏省食品先进制造装备技术重点实验室,江苏无锡214122)
Author(s):
LU Yu1; 2; YU Jing-hu1; 2; ZHU Xing-fei1; 2; ZHANG Bu-fan1; 2
(1.School of Mechanical Engineering, Jiangnan University, Wuxi 214122, China;2.Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment & Technology, Wuxi 214122, China)
关键词:
水稻病害; 组归一化; 激活函数; 深度可分离卷积; 注意力机制
Keywords:
rice disease; group normalization; activation functions; depthwise separable convolution; attention mechanism
分类号:
S511
DOI:
doi:10.3969/j.issn.1000-4440.2024.02.013
摘要:
水稻病害一直是影响水稻产量的重要因素之一,为了快速、准确地检测水稻病害,本研究提出了一种基于卷积神经网络的轻量级水稻叶片病害识别模型。首先,从参数量的角度对注意力机制进行改进,得到轻量级注意力机制模块,对水稻叶片病害特征图中的潜在注意力信息进行深度挖掘；其次,使用深度可分离卷积代替部分标准卷积,进一步降低模型的参数量；最后,为了提高模型的泛化能力,让模型学习过程更快、更稳定,采用了自带内部归一化属性的扩展型指数线性单元函数(SELU)与外部组归一化模块相结合的方法。通过在公共数据集中进行验证,本研究构建模型的平均精度最高(0.990 0),模型在参数量和平均单次迭代时间方面也有一定优势,与其他模型相比,具有相对较好的性能。
Abstract:
Rice diseases have always been one of the important factors affecting rice yield. In order to quickly and accurately detect rice diseases, this study proposed a lightweight rice leaf disease recognition model based on convolutional neural network. Firstly, from the perspective of the number of parameters, the attention mechanism was improved to obtain a lightweight attention mechanism module, and the potential attention information in the rice leaf disease feature map was deeply mined. Secondly, the depthwise separable convolution was used to replace some standard convolutions to further reduce the parameters of the model. Finally, in order to improve the generalization ability and make the model learning process faster and more stable, a method of combining the scaled exponential linear unit (SELU) activation function with internal normalization attribute and the external group normalization module was adopted. By verifying in the public data set, the average accuracy of the model constructed in this study was the highest (0.990 0). The model also had certain advantages in terms of parameter quantity and average single iteration time. Compared with other models, it had relatively higher performance.

参考文献/References:

[1]许娜,叶桥平. 水稻种植与病虫害防治技术研究[J]. 种子科技,2022,40(21): 46-48.
[2]ES-SAADY Y, MAMMASS I E, YASSA M E, et al. Automatic recognition of plant leaves diseases based on serial combination of two SVM classifiers [C]. Tangiers:IEEE,2016.
[3]GAVHALE K R, GAWANDE U. An overview of the research on plant leaves disease detection using image processing techniques[J]. Iosr Journal of Computer Engineering (IOSR-JCE),2014,16(1):10-16.
[4]WANG G, SUN Y, WANG J. Automatic image-based plant disease severity estimation using deep learning[J]. Computational Intelligence and Neuroscience,2017. DOI:10.1155/2017/2917536.
[5]王献锋,张善文,王震,等. 基于叶片图像和环境信息的黄瓜病害识别方法[J].农业工程学报,2014,30(14):148-153.
[6]ZHANG S W, SHANG Y J, WANG L. Plant disease recognition based on plant leaf image[J]. Journal of Animal and Plant Sciences,2015,25(3):42-45.
[7]XIE C, WANG R, ZHANG J, et al. Multi-level learning features for automatic classification of field crop pests[J]. Computers and Electronics in Agriculture,2018,152:233-241.
[8]SANKARAN S, ASHISH M, REZA E, et al. A review of advanced techniques for detecting plant diseases[J]. Computers and Electronics in Agriculture,2010,72(1):1-13.
[9]LI W, CHEN P, WANG B, et al. Automatic localization and count of agricultural crop pests based on an improved deep learning pipeline[J]. Scientific Reports,2019,9:7024.
[10]孟亮,郭小燕,杜佳举,等. 一种轻量级CNN农作物病害图像识别模型[J]. 江苏农业学报,2021,37(5):1143-1150.
[11]WU Y, HE K. Group normalization[J]. International Journal of Computer Vision,2018,128(3):742-755.
[12]WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]. Munich:ECCV,2018.
[13]HOWARD A G, ZHU M, CHEN B, et al. MobileNets:efficient convolutional neural networks for mobile vision applications[J]. ArXiv,2017. DOI:10.48550/arXiv.1704.04861.
[14]KLAMBAUER G, UNTERTHINER T, MAYR A, et al. Self-normalizing neural networks[J]. Advances in Neural Information Processing Systems,2017,30:972-981.
[15]IOFFE S, SZEGEDY C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]. Lille:PMLR,2015.
[16]王建明,陈响育,杨自忠,等. 不同数据增强方法对模型识别精度的影响[J]. 计算机科学,2022,49(增刊1): 418-423.

备注/Memo

备注/Memo:
收稿日期：2023-04-23基金项目：国家自然科学基金项目(51375209)；江苏省先进食品制造装备与技术重点实验室资助项目(FMZ201901)作者简介：陆煜(1998-),男,安徽滁州人,硕士研究生,主要从事农业工程、深度学习研究。(E-mail)jnluyu@stu.jiangnan.edu.cn通讯作者：俞经虎,(E-mail)jnjxjinghuyu@163.com
更新日期/Last Update: 2024-03-17