Field Weed Identification Method Based on Deep Connection Attention Mechanism

doi:10.3778/j.issn.1002-8331.2108-0077

Abstract

Abstract: In order to achieve fast and accurate recognition of field weed images, a field weed recognition model based on deep connected attention mechanism residual network（DCECA-Resnet50-a） is proposed. Using the residual network as a benchmark, this paper improves the position of residual block downsampling, introduces the attention mechanism and connected attention mechanism modules to better extract the feature information in the images, combines the migration learning strategy to alleviate the overfitting phenomenon caused by small sample data sets, improves the generalization of the model and greatly reduces the training time of the model. The experimental results show that the improved model has the best overall performance and high recognition accuracy, with 96.31% accuracy for weeds and fewer model parameters, and achieves the accurate differentiation of four types of common weeds in pea fields, namely, silverleaf daisy, chaparral, matang and pigweed, which provides a corresponding reference for small sample data in the field of agricultural recognition.

Key words: image recognition, residual networks, attention mechanism, migration learning

摘要： 为了实现田间杂草图像快速、准确识别，提出了一种基于深层连接注意力机制残差网络（DCECA-Resnet50-a）的田间杂草识别模型。以残差网络为基准，改进残差块下采样的位置，同时引入注意力机制和连接注意力机制模块以更好地提取图片中的特征信息，结合迁移学习的策略缓解小样本数据集造成的过拟合现象，提高模型的泛化性并大大减少模型的训练时长。实验结果表明，改进后的模型综合性能最好，有较高的识别准确率，对杂草的识别准确率达到了96.31%且模型参数较少，实现了对银叶菊、小蓬草、马唐和猪殃殃四类豌豆田间常见杂草的准确区分，为农业的小样本数据在识别领域中提供了相应的参考作用。

关键词: 图像识别, 残差网络, 注意力机制, 迁移学习

SHU Yali, ZHANG Guowei, WANG Bo, XU Xiaokang. Field Weed Identification Method Based on Deep Connection Attention Mechanism[J]. Computer Engineering and Applications, 2022, 58(6): 271-277.

疏雅丽, 张国伟, 王博, 徐晓康. 基于深层连接注意力机制的田间杂草识别方法[J]. 计算机工程与应用, 2022, 58(6): 271-277.

References

[1] 倪坤晓，何安华.中国粮食供需形势分析[J].世界农业，2021（2）：10-18.
NI K X，HE A H.Analysis of China’s food supply and demand situation[J].World Agriculture，2021（2）：10-18.
[2] 王敬贤.基于卷积神经网络和迁移学习的农作物病害和杂草图像识别研究[D].合肥：中国科学技术大学，2019.
WANG J X.Research on image identification of crop disease and weed based on CNN and transfer learning[D].Hefei：University of Science and Technology of China，2019.
[3] 王鹏飞.基于深度学习的玉米田间杂草识别技术及应用[D].山东泰安：山东农业大学，2019.
WANG P F.Recognition technology and application of weed corn field based on deep learning[D].Tai’an：Shandong Agricultural University，2019.
[4] LEE J H，KIM Y J，KIM Y W，et al.Spotting malignancies from gastric endoscopic images using deep learning[J].Surgical Endoscopy，2019，89（4）：806-815.
[5] GóMEZ-RíOS A，TABIK S，LUENGO J，et al.Towards highly accurate coral texture images classification using deep convolutional neural networks and data augmentation[J].Expert Systems with Applications，2019，43（2）：118-120.
[6] CLáUDIA B，ANA M，ANTóNIO S.Electrocardiogram beat-classification based on a resnet network[J].Studies in Health Technology and Informatics，2019，30（6）：264-265.
[7] HE K，ZHANG X，REN S，et al.Deep residual learning for image recognition[C]//IEEE Conference on Computer Vision & Pattern Recognition.Las Vegas：IEEE Computer Society，2016.
[8] 左羽，徐文博，吴恋.植物识别融合式双特征卷积神经网络[J].计算机工程与设计，2021，42（6）：1706-1712.
ZUO Y，XU W B，WU L.Hybrid dual-feature convolutional neural network for plant recognition[J].Computer Engineering and Design，2021，42（6）：1706-1712.
[9] 陈加敏，徐杨.注意力金字塔卷积残差网络的表情识别[J/OL].计算机工程与应用：1-11[2021-08-05].http：//kns.cnki.net/kcms/detail/11.2127.TP.20210702.1749.004.html.
CHEN J M，XU Y.Expression recognition based on convolutional residual network of attention pyramid[J/OL].Computer Engineering and Applications：1-11[2021-08-05].http：//kns.cnki.net/kcms/detail/11.2127.TP.20210702.1749.
004.html.
[10] 曾伟辉，李淼，李增，等.基于高阶残差和参数共享反馈卷积神经网络的农作物病害识别[J].电子学报，2019，47（9）：1979-1986.
ZENG W H，LI M，LI Z，et al.High-order residuals and parameter sharing feedback convolutional neural network for crop disease recognition[J].Acta Electronica Sinca，2019，47（9）：1979-1986.
[11] 张宸嘉，朱磊，俞璐.卷积神经网络中的注意力机制综述[J].计算机工程与应用，2021，57（20）：64-72.
ZHANG C J，ZHU L，YU L.Review of attention mechanisms in convolutional neural networks[J].Computer Engineering and Applications，2021，57（20）：64-72.
[12] 段萌，王功鹏，牛常勇.基于卷积神经网络的小样本图像识别方法[J].计算机工程与设计，2018，39（1）：224-229.
DUAN M，WANG G P，NIU C Y.Method of small sample size image recognition based on convolutional neural network[J].Computer Engineering and Design，2018，39（1）：224-229.
[13] 龙满生，欧阳春娟，刘欢，等.基于卷积神经网络与迁移学习的油茶病害识图像识别[J].农业工程学报，2018，34（18）：194-201.
LONG M S，OUYANG C J，LIU H，et al.Image recognition of camellia oleifera diseases based on convolutional neural network and transfer learning[J].Transactions of the Chinese Society of Agricultural Engineering，2018，34（18）：194-201.
[14] 王璨，武新慧，李志伟.基于卷积神经网络提取多尺度分层特征识别玉米杂草[J].农业工程学报，2018，34（5）：144-151.
WANG C，WU X H，LI Z W.Recognition of maize and weed based on multi-scale hierarchical features extracted by convolutional neural network[J].Transactions of the Chinese Society of Agricultural Engineering，2018，34（5）：144-151.
[15] 裴晓芳，张杨.基于改进残差网络的花卉图像分类算法[J].电子器件，2020，43（3）：698-704.
PEI X F，ZHANG Y.Flower image classification algorithm based on improved residual network[J].Chinese Journal of Electron Devices，2020，43（3）：698-704.
[16] HE T，ZHANG Z，ZHANG H，et al.Bag of tricks for image classification with convolutional neural networks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Long Beach，CA，US：IEEE，2019：558-567.
[17] WANG Q，WU B，ZHU P，et al.ECA-Net：efficient channel attention for deep convolutional neural networks[C]//IEEE/CVF Conferenceon Computer Vision and Pattern Recognition（CVPR），Seattle，WA，USA，2020：11531-11539.
[18] HU J，SHEN L，ALBANIE S，et al.Squeeze and excitation networks[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2020，42（8）：2011-2023.
[19] MA X，GUO J D，TANG S H，et al.DCANet：learningconnected attentions for convolutional neural networks[J].arXiv：2007.05099，2020.
[20] WOO S，PARK J，LEE J Y，et al.CBAM：convolutional block attention module[J].arXiv：1807.06521，2018.