A Multiclass Approach for Flower Classification Using Convolutional Neural Network

• Diego Nunes Noceli IF Sudeste MG
• Alessandra Martins Coelho IF Sudeste MG
• Matheus de Freitas Oliveira Baffa IF Sudeste MG

DOI: https://doi.org/10.5753/encompif.2021.15952

Abstract

In Brazil, there are about 30 thousand flowers species, with 250 to 300 new species being discovered annually, which may contain different shapes, colors and perfumes. Due to the similar characteristics found in plants, their classification is not a trivial task. This work aims to develop a preliminary method of flower classification, based on Convolutional Neural Networks to identify patterns and automatically differentiate the types of flowers. For the development, a database of 4,135 images was used, containing five groups of flowers, being the dandelion, the daisy, the rose, the tulip and the sunflower. These images were used in the training and evaluation of two types of Convolutional Neural Networks, LeNet-5 and ResNet152. Although simpler, LeNet-5 obtained a superior result, reaching 99.69 % of categorical accuracy.

Keywords: Computer Vision, Deep Learning, Flower Classification

References

Azevedo, E. and Conci, A. (2003).Computação Gráfica: Teoria e Prática, volume 2. Elsevier.

Backes, A. R. and de Mesquita Sá Junior, J. J. (2019).Introdução à visão computacional usando Matlab. Alta Books Editora.

Chaudhary, R. (2020).Final flowers course project dataset. Disponível em:<https://www.kaggle.com/rishitchs/final-flowers-course-project-dataset/metadata>. Acesso em: 21 nov. 2020.

Fernandes, M. P., Sato, J. R., Busatto Filho, G., and Thomaz, C. E. (2012). Classificação estatística e predição da doença de alzheimer por meio de imagens médicas do encéfalo humano. Avanços da Visão Computacional. Omnipax Editora.

Hiary, H., Saadeh, H., Saadeh, M., and Yaqub, M. (2018). Flower classification using deep convolutional neural networks. IET Computer Vision, 12(6):855–862.

Joly, A., Goëau, H., Bonnet, P., Bakić, V., Barbe, J., Selmi, S., Yahiaoui, I., Carré, J., Mouysset, E., Molino, J.-F., et al. (2014). Interactive plant identification based on social image data. Ecological Informatics, 23:22–34.

Kumar, N., Belhumeur, P. N., Biswas, A., Jacobs, D. W., Kress, W. J., Lopez, I. C., andSoares, J. V. (2012). Leafsnap: A computer vision system for automatic plant species identification. In European Conference on Computer Vision, pages 502–516. Springer.

LeCun, Y., Bengio, Y., et al. (1995). Convolutional networks for images, speech, and timeseries. The handbook of brain theory and neural networks, 3361(10): 1995.

Moen, E., Bannon, D., Kudo, T., Graf, W., Covert, M., and Van Valen, D. (2019). Deep learning for cellular image analysis. Nature methods, pages 1–14.

Neves, L. A. P., VIEIRA NETO, H., and Gonzaga, A., editors (2012). Avanços em Visão Computacional. Omnipax, Curitiba, PR, 1 edition. 406 p.

Nieto, D., Brill, A., Feng, Q., Humensky, T., Kim, B., Miener, T., Mukherjee, R., andSevilla, J. (2019). Ctlearn: Deep learning for gamma-ray astronomy. arXiv preprint arXiv:1912.09877.

Oyama, P. d. C., JORGE, L. d. C., Rodrigues, E. L. L., and Gomes, C. C. (2012). Sistema para classificação automática de café em grãos por cor e forma através de imagens digitais. Embrapa Instrumentação - Capítulo em livro científico (ALICE).

Wu, S. G., Bao, F. S., Xu, E. Y., Wang, Y.-X., Chang, Y.-F., and Xiang, Q.-L. (2007). A leaf recognition algorithm for plant classification using probabilistic neural network. In 2007 IEEE international symposium on signal processing and information technology, pages 11–16. IEEE.

Wu, Z., Shen, C., and Van Den Hengel, A. (2019). Wider or deeper: Revisiting the resnet model for visual recognition. Pattern Recognition, 90:119–133.