Yaprak Sınıflandırmak için Yeni Bir Evrişimli Sinir Ağı Modeli Geliştirilmesi

Abstract

Doğanın bir parçası olan bitkiler çevremize güzellik katmanın yanı sıra alternatif tıp gibi farklı sebep için de kullanılmaktadır. Bu gibi uzmanlık gerektiren durumlarda halk arasında yayılan yanlış bilgilerle zehirli bitkilerin şifalı olduğu düşünülerek kullanılması ölüme kadar gidebilecek sorunlara yol açmaktadır. Bu çalışmada yapay zeka teknikleri kullanılarak yaprak görüntülerindeki yaprak türlerinin belirlendiği bir sistem aracılığıyla bu sorunlara çözüm sağlanması amaçlanmaktadır. Son zamanlarda yaygın olarak kullanılan yapay zeka tekniklerinden biri olan evrişimli sinir ağı kullanılmıştır. Çok katmanlı yapısı, birçok parametreye sahip olması ve çok fazla ön işlem gerektirmeden öznitelik öğrenebilmesi, birçok çalışmada kullanılmasının nedenlerinden biridir. Bu çalışmada, sabit bir arka plana sahip yaprak görüntülerinden oluşan 5 farklı veri seti ile evrişimli sinir ağının eğitimi ayrı ayrı yapılmış ve bu eğitim sonucu parametrelerin eğitime olan etkisi incelenmiştir. Bu veri setlerinin birleştirilmesiyle elde edilen 270 türden oluşan birleştirilmiş bir veri seti oluşturulmuştur. Evrişimli sinir ağı ile genel amaçlı bir yaprak sınıflandırma modeli elde edilmiştir. Sınıflandırma işlemi ile elde edilen sonuçlar literatürdeki çalışmalar ile karşılaştırılmıştır.

Keywords

• Segmentasyon
• Derin Öğrenme
• Görüntü İşleme
• Yaprak Sınıflandırma

Developing a Novel CNN Model for Leaf Classification

Abstract

Plants, which are a part of nature, are used for different reasons, such as alternative medicine as well as adding beauty to our environment. In such cases requiring expertise, the misinformation spread among the public and the use of poisonous plants considering that they are medicinal causes problems that can go up to death. In this study, it is aimed to solve these problems through a system that determines the species of leaves in leaf images using artificial intelligence techniques. Convolutional Neural Network (CNN), one of the most widely used artificial intelligence techniques, has been used recently. Its multi-layer structure, having many parameters and being able to learn features without requiring too much pre-processing is one of the reasons why it is used in many studies. In this study, the training of the convolutional neural network was carried out separately with 5 different data sets consisting of leaf images with a fixed background, and the effect of these training parameters on training was investigated. A combined data set consisting of 270 species obtained by combining these data sets was created. A general purpose leaf classification model is obtained with convolutional neural network. The results obtained by the classification process were compared with the studies in the literature.

Keywords

• Segmentation
• Deep Learning
• Image Processing
• Leaf Classification

References

1. Camgözlü, Y. & Kutlu, Y. (2019). Analysis of Pooling Effect on CNN using Leaf Database. Natural and Engineering Sciences, 4(3), 118 – 123.
2. Camgözlü, Y. & Kutlu, Y. (2020). Analysis of Filter Size Effect in Deep Learning. arXiv: 2101.01115.
3. Camgözlü, Y. & Kutlu, Y. (2020). Derin Öğrenme ile Yaprak Sınıflandırma da Görüntü Boyutu Arka Plan Rengi ve Gri Resim ile Renkli Resim Arasındaki Farkın İncelenmesi.Akıllı Sistemler ve Uygulamaları Dergisi, 3(2), 130-133.
4. Tsolakidis, D., Kosmopoulos, D. & Papadourakis, G. (2014). Plant Leaf Recognition Using Zernike Moments and Histogram of Oriented Gradients. Artificial Intelligence: Methods and Applications. SETN 2014. Lecture Notes in Computer Science, 8445, 406-417.
5. Tomar, D. & Agarwal, S. (2016).Leaf Recognition for Plant Classification Using Direct Acyclic Graph Based Multi-Class Least Squares Twin Support Vector Machines. International Journal of Image and Graphic, 16 (3), 1650012-1 – 1650012-17.
6. Wang, Z., Sun, X., Ma, Y., Zhang, H., Ma, Y., & Xie, W. (2014). Plant Recognition Based on Intersecting Cortical Model. 2014 International Joint Conference on Neural Networks, 975-980.
7. Wang, X., Du, W., Guo, F. & Hu, S. (2020). Leaf Recognition Based on Elliptical Half Gabor and Maximum Gap Local Line Direction Pattern. IEEE Access, 8, 39175 – 39183.
8. Zhang, Y., Cui, J., Wang, Z., Kang, J. & Min, Y. (2020).Leaf Image Recognition Based on Bag of Features. Applied Sciences, 10, 5177 - 5194.

9. Kır, B., Öz, C. & Gülbağ, A. (2012). K-NN Sınıflandırma Algoritması Kullanılarak Yaprak Tanıma. 20. Signal Processing and Communications Applications Conference, 18 – 20 April 2012, Fethiye, Muğla Turkey, 1 - 4.
10. Lavania, S. & Matey, P. (2014). Leaf Recognition using Contour Based Edge Detection and SIFT Algorithm. 2014 IEEE International Conference on Computational Intelligence and Computing Research, 1-4.
11. Keivani, M., Mazloum, J., Sedaghatfar, E. & Tavakoli, M. (2020).Automated Analysis of Leaf Shape, Texture and Color Features for Plant Classification. International Information and Engineering Technology Association, Traitement du Signal, 37(1), 17 – 28.
12. Wu, S., Bao, F., Xu, E., Wang, Y., Chang, Y. & Xiang, Q. (2007). A Leaf Recognition Algorithm for Plant Classification Using Probabilistic Neural Network. 2007 IEEE Int. Symp. Signal Process. Inf. Technol. 11-16.
13. Kadir, A., Edi, L., Susanto, A. & Santosa, P. (2011). Leaf Classification Using Shape, Color and Texture Features. International Journal of Computer Trends and Technology, 1(3), 306-311.
14. Lee, S., Chan, C., Mayo, S. & Remagnino, P. (2017). How Deep Learning Extract and Learns Leaf Features for Plant Classification. Pattern Recognition, 71, 1-13.
15. Wang, Z., Sun, X., Yang, Z., Zhang, Y., Ying, Z. & Ma, Y. (2017).Leaf Recognition Based on DPCNN and BOW. Neural Processing Letters, 47, 99 – 115.
16. Hewitt, C. & Mahmoud, M. (2018).Shape-only Features for Plant Leaf Identification. arXiv:1811.08398.
17. Beikmohammadi, A. & Faez, K. (2018). Leaf Classification for Plant Recognition with Deep Transfer Learning. 2018 4th Iranian Conference on Signal Processing and Systems, 21-26.
18. Barre, P., Stöver, B., Müller, K. & Steinhage, V. (2017). Leaf Net: A Computer Vision System for Automatic Plant Species Identification. Ecological Informatics, 40, 50 – 56.
19. Krause, J., Baek, K., Lim, L. & Sugita, G. (2018). WTPlant (What’s That Plant?): A Deep Learning System for Identification Plants in Natural Images. International Conference on Multimedia Retrieval, 517-520.
20. Atabay, H. (2016).A Convolutional Neural Network with A New Architecture on Leaf Classification. IIOAB, 7(5), 326 – 331.
21. Shah, M., Singha, S. & Awate, S. (2017). Leaf Classification using Marginalized Shape Context and Shape + Texture Dual-Path Deep Convolutional Neural Network. 2017 International Conference on Image Processing, 860-864.
22. Chouhan S., Singh, U., Kaul, A. & Jain, S. (2019). A Data Repository of Leaf Images: Practice towards Plant Conservation with Plant Pathology. 2019 4th International Conference on Information Systems and Computer Networks (ISCON), 700-707.
23. Söderkvist, O. (2001). Computer Vision of Leaves from Swedish Trees. Master’s thesis, Linkoping University, The Institute of Technology, Department of Electrical Engineering, Computer Vision.
24. Silva, P., Marcal, A. & Silva, R. (2013). Evaluation of Features for Leaf Discrimination. Springer Lecture Notes in Computer Science, 7950, 197-204.
25. Kumar, N., Belhumeur, P., Biswas, A., Jacobs, D., Kress, W., Lopez, I. & Soares, V. (2012). Leafsnap: A Computer Vision System for Automatic Plant Species Identification. European Conference on Computer Vision (ECCV 2012), 502-516.
26. Humphery, E. & Bello, J. (2012).Rethinking Automatic Chord Recognition with Convolutional Neural Networks. 2012 11th International Conference on Machine Learning and Applications, 357-362.