Fungus Classification Based on CNN Deep Learning Model

Abstract

Artificial intelligence has been developing day by day and has started to take a more prominent place in human life. As computer technologies advance, research on artificial intelligence has also increased in this direction. One of the main goals of this research is to examine how real problems in human life can be solved using artificial intelligence-based deep learning, and to present a case study. Poisoning from the consumption of poisonous fungi is a common problem worldwide. To prevent these poisonings, a mobile application has been developed using Convolutional Neural Networks (CNNs) and transfer learning to detect the species of fungus. The application informs the user about the type of fungus, whether it is poisonous or non-toxic, and whether it is safe to eat. The aim of this study is to reduce poisoning events caused by incorrect fungus detection and to facilitate the identification of fungus species. The developed deep learning model is integrated into a mobile application developed by Flutter that is a mobile application development framework, which enable the detection of fungus species from images taken from the camera or selected from the gallery. CNNs and the EfficientNetV2 model, a transfer learning method, were used. By using these two methods together, the classification accuracy rate for 77 fungus species was obtained as 97%.

Keywords

• Deep Learning
• Convolutional Neural Networks
• Flutter
• Fungus Classification
• Image Classification

References

1. [1] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, no. 7553, pp. 436–444, 2015.
2. [2] K. O’Shea and R. Nash, “An Introduction to Convolutional Neural Networks,” arXiv [cs.NE], 2015.
3. [3] F. Sultana, A. Sufian, and P. Dutta, “Advancements in image classification using convolutional Neural Network,” arXiv [cs.CV], 2019.
4. [4] L. Picek, M. Šulc, J. Matas, J. Heilmann-Clausen, T. S. Jeppesen, and E. Lind, “Automatic fungi recognition: Deep learning meets mycology,” Sensors (Basel), vol. 22, no. 2, p. 633, 2022.
5. [5] S. Sladojevic et al., Fungi Recognition: A Practical Use Case. 2020.
6. [6] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” Commun. ACM, vol. 60, no. 6, pp. 84–90, 2017.
7. [7] K. He, X. Zhang, S. Ren, and J. Sun, “Delving deep into rectifiers: Surpassing human-level performance on ImageNet classification,” in 2015 IEEE International Conference on Computer Vision (ICCV), 2015.
8. [8] K. Kamnitsas et al., “Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation,” Med. Image Anal., vol. 36, pp. 61–78, 2017.

9. [9] N. Z. Kayalı and S. Ve Ilhan Omurca, Konvolüsyonel Sinir Ağları (CNN) ile Çin Sayı Örüntülerinin Sınıflandırması. 2021.
10. [10] F. Bozkurt ve M. Yağanoğlu, "Derin Evrişimli Sinir Ağları Kullanarak Akciğer X-Ray Görüntülerinden COVID-19 Tespiti", Veri Bilimi, c. 4, sayı. 2, ss. 1-8, Ağu. 2021.
11. [11] İ. Ökten ve U. Yüzgeç, "Evrişimli Sinir Ağı ile Çeltik Bitkisi Hastalığının Tespiti", Bitlis Eren Üniversitesi Fen Bilimleri Dergisi, c. 11, sayı. 1, ss. 203-217, Mar. 2022, doi:10.17798/bitlisfen.1014393.
12. [12] “What are Convolutional Neural Networks?” Ibm.com. [Online]. Available: https://www.ibm.com/topics/convolutional-neural-networks. [Accessed: 27-Dec-2022].
13. [13] A. H. Reynolds, “Anh H. reynolds,” Anh H. Reynolds. [Online]. Available: https://anhreynolds.com/blogs/cnn.html. [Accessed: 27-Dec-2022].
14. [14] A. Kızrak, “DERİNE DAHA DERİNE: Evrişimli Sinir Ağları - ayyüce kızrak, ph.D,” Medium, 28-May-2018. [Online]. Available: https://ayyucekizrak.medium.com/deri%CC%87ne-daha-deri%CC%87ne-evri%C5%9Fimli-sinir-a%C4%9Flar%C4%B1-2813a2c8b2a9. [Accessed: 27-Dec-2022].
15. [15] S. Saha, “A comprehensive guide to convolutional neural networks — the ELI5 way,” Towards Data Science, 15-Dec-2018. [Online]. Available: https://towardsdatascience.com/a-comprehensive-guide-to-convolutional-neural-networks-the-eli5-way-3bd2b1164a53. [Accessed: 27-Dec-2022].
16. [16] M. Hussain, J. J. Bird, and D. R. Faria, “A study on CNN transfer learning for image classification,” in Advances in Intelligent Systems and Computing, Cham: Springer International Publishing, 2019, pp. 191–202.
17. [17] H.-C. Shin et al., “Deep convolutional neural networks for computer-aided detection: CNN architectures, dataset characteristics and transfer learning,” IEEE Trans. Med. Imaging, vol. 35, no. 5, pp. 1285–1298, 2016.
18. [18] S. Lee, “(TF2) Transfer Learning - Feature Extraction,” AAA (All About AI), 05-Mar-2022. [Online]. Available: https://seunghan96.github.io/dlf/TF2_4%EC%9E%A5/. [Accessed: 27-Dec-2022].
19. [19] M. Tan and Q. V. Le, “EfficientNetV2: Smaller models and faster training,” arXiv [cs.CV], 2021.
20. [20] Wikipedia contributors, “Flutter,” Wikipedia, The Free Encyclopedia. [Online]. Available: https://tr.wikipedia.org/w/index.php?title=Flutter&oldid=27787028.
21. [21] “2018 FGCVx fungi classification challenge,” Kaggle.com. [Online]. Available: https://www.kaggle.com/competitions/fungi-challenge-fgvc-2018/overview. [Accessed: 27-Dec-2022].
22. [22] “Danmarks officielle database for svampefund,” Danmarks SvampeatlasXXXXX. [Online]. Available: https://svampe.databasen.org/en/. [Accessed: 27-Dec-2022].
23. [23] Z. Zhong, L. Zheng, G. Kang, S. Li, and Y. Yang, “Random Erasing Data Augmentation,” Proc. Conf. AAAI Artif. Intell., vol. 34, no. 07, pp. 13001–13008, 2020.
24. [24] W. Li, C. Chen, M. Zhang, H. Li, and Q. Du, “Data augmentation for hyperspectral image classification with deep CNN,” IEEE Geosci. Remote Sens. Lett., vol. 16, no. 4, pp. 593–597, 2019.