Abstract
Tobacco classification is a challenging research topic and plays a crucial role in the process of cigarette production. Tobacco classification mainly relies on manual selection, which is time-consuming, labor-intensive, and subjective. With the development of machine learning, how to automatically classify tobacco leaves has become a fast-growing research issue. However, the lack of high-quality tobacco image dataset limits the applications of deep learning approaches seriously. In addition, a large number of parameters, single-scale features, and high computational complexity further affect the classification results. To address these problems, we establish a new tobacco dataset, which contains 11849 labeled tobacco leaf images in 9 categories. Then, a multi-channel and multi-scale separable dilated convolution neural network with attention mechanism is proposed. The adopted separable dilated convolution increases the receptive fields of the convolution kernels and improves the calculation speed and accuracy of the model without increasing the number of training parameters. Then, the attention mechanism is integrated, so that the network can assign higher weights for discriminative features and get rid of redundant ones by assigning lower even zero weights. Experimental results demonstrate the superiority and effectiveness of the proposed approach. Based on our new dataset, the average classification result of the proposed method achieves 98.4%.
Similar content being viewed by others
Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
References
Dasari SK, Prasad V (2019) A novel and proposed comprehensive methodology using deep convolutional neural networks for flue cured tobacco leaves classification. Int J Inf Technol 11(1):107–117
Liu HB, He LY, Ma WJ (2007) Color features of transmittance images applied in the classification of tobacco leaves. Trans CSAE 23(9):169–171
Li H, Zhao HD, Shen JY (2015) Screening the effective features in the near-infrared spectroscopy of tobacco leaf based on BPSO and SVM. Phys Exp 35(6):8–12
Zhang Y, He LY (2011) Auto-grouping method of flue-cured tobacco leaves based on near infrared spectra technology. Trans CSAE 27(4):350–354
Zhang XH, Zhang F (2008) Images features extraction of tobacco leaves. Congr Imag Signal Process 2:773–776
Pei LQ, Shen JY, Liu RJ (2017) Deep feature of image screened by improved clustering algorithm cascaded with genetic algorithm. In: 29th Chinese control and decision conference 452-455
Ma XL, Shen JY, Liu RJ (2016) Choice of tobacco leaf features based on selected probability of particle swarm algorithm. In: 28th Chinese control and decision conference 3041-3044
Liu JJ, Shen JY, Shen Z (2013) Grading tobacco leaves based on image processing and generalized regression neural network. Trans Chin Soc Agri Mach 44(11):287–292
Pouyanfar S, Sadiq S, Yan YL (2019) A survey on deep learning: algorithms, techniques, and applications. ACM Comput Surv 51(5):92
Liu JH, Guo F, Gao H, Huang ZG (2021) Image classification method on class imbalance datasets using multi-scale CNN and two-stage transfer learning. Neural Comput Appl 33(21):14179–14197
Shervin M, Yuri B, Fatih P (2022) Image segmentation using deep learning: a survey. IEEE Trans Pattern Anal Mach Intell 44(7):3523–3542
Ke RH, Bugeau A, Papadakis N (2021) Multi-task deep learning for image segmentation using recursive approximation tasks. IEEE Trans Imag Process 30:3555–3567
Wen LW, Ding JS, Xu Z (2022) Multiframe detection of sea-surface small target using deep convolutional neural network. IEEE Trans Geosci Remote Sens 60:5107116
Bukhsh ZA, Jansen N, Saeed A (2021) Damage detection using in-domain and cross-domain transfer learning. Neural Comput Appl 33(24):16921–16936
Zhang HT, Tian M, Shao GP (2022) Target detection of forward-looking sonar image based on improved YOLOv5. IEEE Acesss 10:18023–18034
He L, Amy R, Aaron C (2021) Spatial segmentation for processing videos for farming automation. Comput Electron Agri 184:106095
Sergiu O, Pablo M, Alberto G, John AC (2022) A review on deep learning techniques for video prediction. IEEE Trans Pattern Anal Mach Intell 44(6):2806–2862
Li A, Zheng BY, Li L, Zhang C (2020) Optical flow estimation and denoising of video images based on deep learning models. IEEE Access 8:144122–144135
Jiao LC, Zhang RH, Liu F (2022) New generation deep learning for video object detection: a survey. IEEE Trans Neural Netw Learn Syst 3195-3215
Che A, Muhamad F, Suandi Shahrel A (2022) Video surveillance image enhancement via a convolutional neural network and stacked denoising autoencoder. Neural Comput Appl 34(4):3079–3095
Huan JL, Sekh A, Quek C, Prasad D (2022) Emotionally charged text classification with deep learning and sentiment semantic. Neural Comput Appl 34(3):2353–2374
Wang SQ, Chen Z, You SR, Wang BC (2022) Brain stroke lesion segmentation using consistent perception generative adversarial network. Neural Comput Appl 34(11):8657–8669
Khan S, Islam N, Jan Z (2019) A novel deep learning based framework for the detection and classification of breast cancer using transfer learning. Pattern Recogn Lett 125:1–6
Hanene BY, Ben C, Majid S (2022) Multitask deep learning reconstruction and localization of lesions in limited angle diffuse optical tomography. IEEE Trans Medi Imag 41(3):515–530
Fan CH, Yi JY, Tao JH (2020) Gated recurrent fusion with joint training framework for robust end-to-end speech recognition. IEEE/ACM Trans Audio, Speech Lang Process 29:198–209
Tu YH, Du J, Lee CH (2019) Speech enhancement based on teacher-student deep learning using improved speech presence probability for noise-robust speech recognition. IEEE/ACM Trans Audio Speech Lang Process 27(12):2080–2091
Yesim D, Zekeriya T (2021) Mini-batch sample selection strategies for deep learning based speech recognition. Appl Acoust 171:107573
Siva KD, Vadamodula P (2019) A novel and proposed comprehensive methodology using deep convolutional neural networks for flue cured tobacco leaves classification. Int J Inf Technol 11(1):107–117
Li GC, Zhen HJ, Jiao FY (2021) Research on tobacco leaf grading algorithm based on transfer learning. 2021 IEEE International Conference on Artificial Intelligence and Computer Applications (ICAICA) Dalian, China 2021
Dwiretno IS, Handayani T, Chastine F (2019) Classification of tobacco leaf pests using VGG16 transfer learning. In: 12th International Conference on Information and Communication Technology and System (ICTS) Surabaya, Indonesia 2019
Lu MY, Jiang SW, Wang C, Chen D (2022) Tobacco leaf grading based on deep convolutional neural networks and machine vision. J ASABE 65(1):11–22
Yann L (2015) LeNet-5, convolutional neural networks. [Online]. Available: htttp://yann.lecun.com/exdb/lenet.2015
Krizhevsky A, Sutskever I, Hinton G (2017) ImageNet classification with deep convolutional neural networks. Commun ACM 60(5):84–90
Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. Computer Science
Szegedy C, Liu W, Jia Y, Sermanet P, Reed S (2015) Going deeper with convolutions. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition 7:1–9
Shao SY, McAleer S, Yan RQ (2019) Highly accurate machine fault diagnosis using deep transfer learning. IEEE Trans Ind Inf 15(4):2446–2455
Voravee P, Kanchana S, Krisanarach N (2022) Hybrid differential evolution and particle swarm optimization for multi-visit and multi-period workforce scheduling and routing problems. Comput Electron Agri 197:106929
Vijai S, Varsha (2015) Detection of unhealthy region of plant leaves using image processing and genetic algorithm. 2015 International Conference on Advances in Computer Engineering and Applications
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Xu, M., Gao, J., Zhang, Z. et al. Multi-channel and multi-scale separable dilated convolutional neural network with attention mechanism for flue-cured tobacco classification. Neural Comput & Applic 35, 15511–15529 (2023). https://doi.org/10.1007/s00521-023-08544-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-023-08544-7