Konvolüsyonel yapay sinir ağları ve öğrenme transferi ile bitki tanıma

Öz

Doğa büyük miktarda bitki ve çiçek türü zenginliğine sahiptir ve bu çeşitlilik nedeniyle bu türlerin tanımlanması botanik alanında uzmanlık gerektirmektedir. Bir otomatik bitki tanıma sisteminin geliştirilmesi bu süreci kolaylaştıracaktır. Bu çalışmada, böyle bir sistemi geliştirmek için konvolüsyonel yapay sinir ağlarından ve öğrenme transferinden faydalanılmıştır. Veritabanındaki görüntüler diğer veritabanlarından ve webden toplanmıştır ve toplamda 76 türe ait 5.345 çiçek ve bitkiden oluşmaktadır. Türlerin 65 tanesi çeşitli çiçek türleridir ve 11 tanesi ise diğer bitki çeşitlerindendir. Veritabanındaki görüntü sayısını ve modelin genelleme kapasitesini arttırmak için çeşitli veri çoğaltma yöntemleri uygulanmıştır. Veri çoğaltmak için, 4 açıdan rastgele döndürme, [-0.2, 0.2] aralığında rastgele parlaklık değişimi ve yatay yansıtma işlemleri uygulanmıştır. Aynı zamanda, görüntüleri modele girdi olarak vermeden önce, görüntülere merkezden kesme ve normalizasyon işlemleri uygulanmıştır. Geliştirilen model eğitim verileri için 0.9971, test verileri için ise 0.9897 isabet oranı elde etmiştir.

Anahtar Kelimeler

• Yapay zekâ
• Derin öğrenme
• Konvolüsyonel yapay sinir ağları
• Tanımlama
• Sınıflandırma

Plant identification with convolutional neural networks and transfer learning

Abstract

Nature is rich with a vast amount of plant and flower species and because of their great diversity; identification of these species requires expertise in the field. Development of an automatic plant identification system can ease this process. In this work, deep Convolutional Neural Networks and Transfer Learning have been utilized in order to develop such an identification system. Images in the database have been collected from other databases and the web and in total it consists of 5,345 flowers and plant images belong to 76 species. 65 of the species are various flower species and 11 of them are other plant species. Data augmentation techniques has been applied in order to increase the number of images in the database and to improve the generalization capacity of the model. For data augmentation, random rotation at four angles, random brightness change in the range of [-0.2, 0.2] and horizontal flip have been applied. Also preprocessing techniques such as center cropping and normalizing have been applied to images before input them to the model. In automatic plant recognition, 0.9971 accuracy achieved on the training set and 0.9897 accuracy achieved on the test set

Keywords

• Artificial intelligence
• Deep learning
• Convolutional neural networks
• Identification
• Classification

References

1. [1] O’Shea K et al. “An Introduction to Convolutional Neural Networks”. https://arxiv.org/abs/1511.08458 (31.03.2021).
2. [2] ImageNet. “ImageNet Large Scale Visual Recognition Challenge”. http://image-net.org/challenges/LSVRC/ (01.05.2020).
3. [3] Zeiler MD, Fergus R. “Visualizing and understanding convolutional networks”. ECCV 2014. Zurich, Switzerland, 6-12 September 2014.
4. [4] Goodfellow I, Bengio Y, Courville A. Transfer Learning. Editors: Dietterich T. Deep Learning, 526-531, Cambridge, Massachusetts, USA, MIT Press, 2016.
5. [5] Shorten C, Khoshgoftaar TM. “A survey on Image Data Augmentation for deep learning”. Journal of Big Data, 2019. https://doi.org/10.1186/s40537-019-0197-0.
6. [6] Lohr S. Sampling: Design and Analysis. Editors: Blitzstein SK, Faraway JJ, Tanner M, Zidek J. Stratified Sampling, 73-115, Boca Raton, FL, USA, CRC Press, 2012.
7. [7] Krizhevsky A, Sutskever I, Hinton GE. “Imagenet classification with deep convolutional neural networks”. Neural Information Processing Systems. Lake Tahoe, Nevada, USA, 3-6 December 2012.
8. [8] Khan A, Sohail A, Zahoora U, Qureshi AS. “A Survey of the recent architectures of deep convolutional neural networks”. Artficial Intelligence Review, 2020. https://doi.org/10.1007/s10462-020-09825-6.

9. [9] Tan C, Sun F, Kong T, Zhang W, Yang C, Liu C. “A survey on deep transfer learning”. International Conference on Artficial Neural Networks. Rhodes, Greece, 4-7 October 2018.
10. [10] Wang J, Perez L. ‘‘The effectiveness of data augmentation in ımage classification using deep learning’’. https://arxiv.org/abs/1712.04621 (31.03.2021).
11. [11] Dyrmann M, Karstoft H, Midtiby HS.” Plant species classification using deep convolutional neural network”. Biosystems Engineering, 151, 72-80, 2016.
12. [12] Grinblat GL, Uzal LC, Larese MG. “Deep learning for plant identification using vein morphological patterns”. Computers and Electronics in Agriculture, 27, 418-424, 2016.
13. [13] Kaya A, Keçeli AS, Çatal Ç, Yalic HY, Huseyin T, Tekinerdoğan B. “Analysis of transfer learning for deep neural network based plant classification models”. Computers and Electronics in Agriculture, 158, 20-29, 2019.
14. [14] Lee HS, Chan CS, Mayo S, Remagnino P. “How deep learning extracts and learns leaf features for plant classification”. Pattern Recognition, 71, 1-13, 2017.
15. [15] Lasseck M. “Image-Based plant species ıdentification with deep convolutional neural networks”. CLEF, Dublin, Ireland, 11-14 September 2017.
16. [16] Ghazi MM, Yanikoğlu B, Aptoula E. “Plant identification using deep neural networks via optimization of transfer learning parameters”. Neurocomputing, 235, 228-235, 2017.
17. [17] Šulc M, Picek L, Matas J. “Plant recognition by inception networks with test-time class prior estimation”. CLEF, Avignon, France, 10-14 September 2018.
18. [18] University of Oxford. “102 Category Flower Dataset”. http://www.robots.ox.ac.uk/~vgg/data/flowers/102/in dex.html, (15.06.2019).
19. [19] Github. “Tolga-Karahan/Mobile-Plant-Identification”. https://tolga-karahan.github.io/Mobile-PlantIdentification/(13.11.2020).
20. [20] Atatürk Orman Çiftliği. “Atatürk Orman Çiftliği”. http://www.aoc.gov.tr, (08.09.2020).
21. [21] Yabani Çiçekler. “Yaban Çiçekler”. www.yabanicicekler.com, (08.09.2020).
22. [22] Yosinski J, Clune J, Bengio Y, Lipson H. “How Transferable are Features in Deep Neural Networks?”. https://arxiv.org/abs/1411.1792 (31.03.2021).
23. [23] Howard AG et al. “MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications”. https://arxiv.org/abs/1704.04861 (31.03.2021).
24. [24] Wikipedia Commons. “File: Typical cnn.png”. https://commons.wikimedia.org/wiki/File:Typical_cnn.p ng (12.11.2020).
25. [25] Machine Think. “MobileNet Version 2” https://machinethink.net/blog/mobilenet-v2/ (05.05.2020).
26. [26] Sandler M, Howard A, Zhu M, Zhmoginov A, Chen L. "Mobilenetv2: Inverted residuals and linear bottlenecks". CVPR, Salt Lake City, USA, 19-21 June 2018.
27. [27] Kingma DP, Ba JL. “Adam: A method for stochastic optimization”. ICLR, San Diego, USA,7-9 May 2015.
28. [28] Kellejer JD, Mac Namee B, D’arcy A. Fundamentals of Machine Learning for Predictive Data Analytics. Editor: Mellon M. Precision, Recall, and F1 Measure, 413-417, Cambridge, Massachusetts, USA, The MIT Press, 2015.