基于多尺度卷积和注意力机制的枣品种识别

doi:10.13733/j.jcam.issn.2095-5553.2024.06.021

摘要/Abstract

摘要：

为提高自然场景下枣品种识别方法的准确率，提出一种融合多尺度卷积和注意力机制的枣品种识别模型(Jujube-ResNet-18)。以自然场景下的10类枣品种为对象，根据枣品种图像的特点，该模型在ResNet-18基础上进行改进。引入多尺度卷积模块，增强模型对枣果多尺度特征的提取能力；在每个残差块中加入注意力机制CBAM，提高枣果特征信息权重，减弱复杂背景等无用特征的影响。试验结果表明，Jujube-ResNet-18在枣品种数据集上的准确率为89.5%，参数量和权重大小分别为1.135×107和43.41 MB。与其他算法相比，Jujube-ResNet-18有更好的特征提取能力、抗干扰能力和较小的模型复杂度，可为自然场景下的枣品种识别研究提供参考。

关键词: 枣品种识别, 深度学习, 残差网络, 多尺度卷积, 注意力机制

Abstract:

In order to improve the accuracy of jujube varieties recognition method in natural scenes, a jujube varieties recognition model (Jujube-ResNet-18) was proposed by integrating multi-scale convolution and attention mechanism. In this study, ten types of jujube varieties under natural scenes were taken as objects. According to the characteristics of jujube variety images, the model in this paper was improved on the basis of ResNet-18. Firstly, the multi-scale convolution module was introduced to enhance the ability of the model to extract multiscale features of jujube fruit. Secondly, the attentional mechanism CBAM was added into each residual block to increase the weight of jujube fruit feature information and weaken the influence of complex background and other useless features. The experimental results showed that the accuracy of Jujube-ResNet-18 on the date variety dataset was 89.5%, while the number of parameters and weight were only 1.135×107 and 43.41 MB, respectively. Compared with other algorithms, Jujube-ResNet-18 has better feature extraction ability, antiinterference ability and smaller model complexity, which can provide a reference for the study of jujube varieties recognition in natural scenes.

Key words: jujube varieties recognition, deep learning, residual network, multi-scale convolution, attention mechanism

中图分类号:

TP391.4

雷浩, 苑迎春, 何振学. 基于多尺度卷积和注意力机制的枣品种识别[J]. 中国农机化学报, 2024, 45(6): 135-141.

Lei Hao, Yuan Yingchun, He Zhenxue. Jujube varieties recognition based on multiscale convolution and attention mechanism[J]. Journal of Chinese Agricultural Mechanization, 2024, 45(6): 135-141.

参考文献

［1］　刘孟军, 王玖瑞, 刘平, 等. 中国枣生产与科研成就及前沿进展［J］. 园艺学报, 2015, 42(9): 1683-1698.Liu Mengjun, Wang Jiurui, Liu Ping, et al. Historical achievements and frontier advances in the production and research of Chinese jujube (ziziphus jujuba) in China ［J］. Acta Horticulturae Sinica, 2015, 42(9): 1683-1698.

［2］　刘孟军, 王玖瑞. 新中国果树科学研究70年——枣［J］. 果树学报, 2019, 36(10): 1369-1381.Liu Mengjun, Wang Jiurui. Fruit scientific research in New China in the past 70 years: Chinese jujube ［J］. Journal of Fruit Science, 2019, 36(10): 1369-1381.

［3］　阳灵燕, 张红燕, 陈玉峰, 等. 机器学习在农作物品种识别中的应用研究进展［J］. 中国农学通报, 2020, 36(30): 158-164.

Yang Lingyan, Zhang Hongyan, Chen Yufeng, et al. The application of machine learning in crop variety recognition: A review ［J］. Chinese Agricultural Science Bulletin, 2020, 36(30): 158-164.

［4］　苏军, 饶元, 张敬尧, 等. 基于GA优化SVM的干制红枣品种分类方法［J］. 洛阳理工学院学报(自然科学版), 2018, 28(4): 65-69, 93.

Su Jun, Rao Yuan, Zhang Jingyao, et al. Study on classification method of jujube varieties based on GA optimized SVM ［J］. Journal of Luoyang Institute of Science and Technology (Natural Science Edition), 2018, 28(4): 65-69, 93.

［5］　Sabanci K. Different apple varieties classification using KNN and MLP algorithms ［J］. International Journal of Intelligent Systems and Applications in Engineering, 2016, 4(S1): 166-169.

［6］　张珂, 冯晓晗, 郭玉荣, 等. 图像分类的深度卷积神经网络模型综述［J］. 中国图象图形学报, 2021, 26(10): 2305-2325.

Zhang Ke, Feng Xiaohan, Guo Yurong, et al. Overview of deep convolutional neural networks for image classification ［J］. Journal of Image and Graphics, 2021, 26(10): 2305-2325.

［7］　Krizhevsky A, Sutskever I, Hinton G E. Image Net classification with deep convolutional neural networks ［J］. Communications of the ACM, 2017, 60(6): 84-90.

［8］　Simonyan K, Zisserman A. Very deep convolutional networks for largescale image recognition ［J］. arXiv Preprint arXiv: 1409.1556, 2014.

［9］　He K, Zhang X, Ren S, et al. Deep residual learning for image recognition ［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 770-778.

［10］　Szegedy C, Liu W, Jia Y, et al. Going deeper with convolutions ［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2015: 1-9.

［11］　余游江, 喻彩丽, 尚远航, 等. 基于Stacking模型的红枣品种分类识别［J］. 北方园艺, 2022(8): 139-148.Yu Youjiang, Yu Caili, Shang Yuanhang, et al. Classification and recognition of jujube varieties based on Stacking model fusion ［J］. Northern Horticulture, 2022(8): 139-148.

［12］　ElKahlout M I, AbuNaser S S. Peach type classification using deep learning ［J］. International Journal of Academic Engineering Research, 2019, 3(12).

［13］　倪建功, 杨昊岩, 李娟, 等. 基于改进型AlexNet的花生荚果品种识别［J］. 花生学报, 2021, 50(4): 14-22.

Ni Jiangong, Yang Haoyan, Li Juan, et al. Variety identification of peanut pod based on improved AlexNet ［J］.Journal of Peanut Science, 2021, 50(4): 14-22.

［14］　赵腾飞, 胡国玉, 周建平, 等. 卷积神经网络算法在核桃仁分类中的研究［J］. 中国农机化学报, 2022, 43(6): 181-189.

Zhao Tengfei, Hu Guoyu, Zhou Jianping, et al. Research on convolutional neural network algorithm for walnut kernel classification identification ［J］. Journal of Chinese Agricultural Mechanization, 2022, 43(6): 181-189.

［15］　Hu J, Shen L, Sun G. Squeeze and excitation networks ［C］. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 7132-7141.

［16］　苏宝峰, 沈磊, 陈山, 等. 基于注意力机制的葡萄品种多特征分类方法［J］. 农业机械学报, 2021, 52(11): 226-233, 252.

Su Baofeng, Shen Lei, Chen Shan, et al. Multifeatures identification of grape cultivars based on attention mechanism ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2021, 52(11): 226-233, 232.

［17］　张继成, 李德顺. 基于深度残差学习的成熟草莓识别方法［J］. 中国农机化学报, 2022, 43(2): 136-142.

Zhang Jicheng, Li Deshun. Ripe strawberry recognition method based on deep residual learning ［J］. Journal of Chinese Agricultural Mechanization, 2022, 43(2): 136-142.

［18］　耿磊, 黄亚龙, 郭永敏. 基于融合注意力机制的苹果品种分类方法［J］. 农业机械学报, 2022, 53(6): 304-310, 369.
Geng Lei, Huang Yalong, Guo Yongmin. Apple variety classification method based on fusion attention mechanism ［J］. Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(6): 304-310, 369.
［19］　马岽奡, 唐娉, 赵理君, 等. 深度学习图像数据增广方法研究综述［J］. 中国图象图形学报, 2021, 26(3): 487-502.
Ma Donggao, Tang Ping, Zhao Lijun, et al. Review of data augmentation for image in deep learning ［J］. Journal of Image and Graphics, 26(3): 487-502.
［20］　Peng G, Tuo X, Shen T, et al. Recognition of rock microfracture signal based on deep convolution neural network inception algorithm ［J］. IEEE Access, 2021, 9: 89390-89399.
［21］　黄英来, 艾昕. 改进残差网络在玉米叶片病害图像的分类研究［J］. 计算机工程与应用, 2021, 57(23): 178-184.
Huang Yinglai, Ai Xin. Research on classification of corn leaf disease image by improved residual network ［J］. Computer Engineering and Applications, 2021, 57(23): 178-184.
［22］　Woo S, Park J, Lee JY, et al. CBAM: Convolutional block attention module ［C］. Proceedings of the European Conference on Computer Vision (ECCV), 2018: 3-19.
［23］　Selvaraju R R, Cogswell M, Das A, et al. Grad-CAM: Visual explanations from deep networks via gradient-based localization ［C］. Proceedings of the IEEE International Conference on Computer Vision, 2017: 618-626.

[1]	岳耀华, 张伟, 亓立强, . 基于改进YOLOv5的田间大豆花朵生长状态检测方法研究[J]. 中国农机化学报, 2024, 45(7): 188-193.
[2]	李运良, 赵明岩, 王鑫, 闫泓硕, 李育蝉. 基于改进ResNet的带孔蛋胚裂纹检测方法#br#[J]. 中国农机化学报, 2024, 45(7): 201-208.
[3]	吴利刚, 吕媛媛, 周倩, 陈乐, 张梁, 史建华. 基于改进YOLOv5的轻量级黄花成熟检测方法[J]. 中国农机化学报, 2024, 45(7): 235-242.
[4]	张继成, 黄向党. 基于改进YOLOv3的玉米病害识别方法[J]. 中国农机化学报, 2024, 45(7): 269-275.
[5]	崔海朋, 秦朝旭, 马志宇. 基于深度学习的鱼类特征点检测与体征识别方法[J]. 中国农机化学报, 2024, 45(6): 201-207.
[6]	黄尧, 何敬, 付饶, 刘刚, , 林远杨. YOLOv5s-CBAM算法在福寿螺虫卵识别中的应用分析[J]. 中国农机化学报, 2024, 45(6): 223-228.
[7]	周康乔, 刘向阳, 郑特驹. 改进YOLOv5的小目标多类别农田害虫检测算法研究[J]. 中国农机化学报, 2024, 45(6): 235-241.
[8]	陶旭, , 余富强, , 蔡金金, 么炜, , 刘博, . 面向激光雷达点云数据的多结构树种识别[J]. 中国农机化学报, 2024, 45(5): 168-175.
[9]	薛霞, 刘鹏, 周文. 基于改进YOLOv8算法的实时细粒度植物病害检测[J]. 中国农机化学报, 2024, 45(5): 188-194.
[10]	漆海霞, , , 李承杰, 黄桂珍. 基于轻量化YOLOv4的死淘鸡目标检测算法[J]. 中国农机化学报, 2024, 45(5): 195-201.
[11]	罗鑫, 李加强, 何超. 基于改进YOLOv4的澳洲坚果视觉监测方法[J]. 中国农机化学报, 2024, 45(5): 217-222.
[12]	廖舒怀, , 王凯, 宋健, 解福祥, 王名声, , 龚中良. 基于移动机器人的树木识别与冠层信息测量[J]. 中国农机化学报, 2024, 45(5): 232-238.
[13]	王化佳, 顾寄南, 王梦妮, 夏子林. 基于深度学习的茶嫩芽分割与采摘点定位方法研究[J]. 中国农机化学报, 2024, 45(5): 246-252.
[14]	霍正瑞, 孙铁波. 基于改进YOLOv5算法的瓜蒌分级方法[J]. 中国农机化学报, 2024, 45(4): 100-107.
[15]	陈鲁威, , 曾锦, , 袁全春, , 夏烨, , 潘健, 吕晓兰, , . 基于改进DeepLabV3+的梨树冠层分割方法[J]. 中国农机化学报, 2024, 45(4): 155-161.