Uma Abordagem Híbrida para Detecção de Moedas de Real Baseada em Algoritmos de Localização e Redes Neurais Convolucionais
Resumo
A detecção de múltiplas moedas é um problema de Visão Computacional que objetiva encontrar moedas em uma imagem e classificá-las quanto aos seus respectivos valores. Para o Real, a literatura tipicamente considera o uso de técnicas de extração de características seguida de classificadores de Machine Learning, em especial as redes neurais artificiais. Porém, considerando resultados recentes com Deep Learning, houve a motivação de propor uma abordagem híbrida considerando a localização de moedas com técnicas tradicionais de Visão Computacional e classificação com Redes Neurais Convolucionais. Múltiplas arquiteturas foram avaliadas e a solução proposta possui mean average precision superior a 92% com acurácia de classificação maior que 97%, superando outro resultado em 6%.
Palavras-chave:
Visão Computacional, Aprendizado Profundo, Redes Neurais Convolucionais
Referências
Ballard, D. H. (1981). Generalizing the Hough Transform to detect arbitrary shapes. Pattern Recognition, 13(2):111–122.
Cameron, A. C. and Windmeijer, F. A. (1997). An R-squared measure of goodness of fit for some common nonlinear regression models. Journal of Econometrics, 77(2):329–342.
Capece, N., Erra, U., and Ciliberto, A. V. (2016). Implementation of a coin recognition system for mobile devices with deep learning. In 12th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS), pages 186–192, Itália. IEEE.
Capece, N., Erra, U., and Ciliberto, A. V. (2016). Implementation of a coin recognition system for mobile devices with deep learning. In 2016 12th International Conference on SignalImage Technology Internet-Based Systems (SITIS), pages 186–192, Itália. IEEE.
Chollet, F. (2017). Deep Learning with Python. Manning Publications, Estados Unidos, 1 edition. de Freitas, S. S. (2014). Determinação do valor total de moedas em imagens digitais. Mografia (Bacharelado em Sistemas de Informação) – Universidade Federal do Ceará.
Everingham, M., , Gool, L. V., Williams, C. K., Winn, J., and Zisserman, A. (2010). The Pascal Visual Object Classes (VOC) challenge. International journal of computer vision, 88(2):303–338.
Farooque, G., Sargano, A. B., Shafi, I., and Ali, W. (2016). Coin recognition with reduced feature set sift algorithm using neural network. In 2016 International Conference on Frontiers of Information Technology (FIT), pages 93–98, Paquistão. IEEE.
Freeman, I., Roese-Koerner, L., and Kummert, A. (2018). Effnet: An efficient structure for convolutional neural networks. In 2018 25th IEEE International Conference on Image Processing (ICIP), pages 6–10. IEEE.
Gupta, S., Kaur, G., Gupta, D., and Jindal, U. (2019). Brazilian coins recognition using histogram of oriented gradients features. Journal of Computational and Theoretical Nanoscience, 16(10):4170–4178.
Iandola, F. N., Han, S., Moskewicz, M. W., Ashraf, K., Dally, W. J., and Keutzer, K. (2016). Squeezenet: Alexnet-level accuracy with 50× fewer parameters and [ 0.5 mb model size. Disponı́vel em https://arxiv.org/abs/1602.07360. Acesso em 24 de agosto de 2020.
Jain, N. and Jain, N. (2012). Coin recognition using circular Hough transform. International Journal of Electronics Communication and Computer Technology, 2(3):101–104.
Jähne, B. and Haußecker, H. (2000). Computer Vision and Applications – A Guide for Students and Practitioners. Academic Press, Estados Unidos, 1 edition.
Khan, S., Rahmani, H., Shah, S. A. A., and Bennamoun, M. (2018). A Guide to Convolutional Neural Networks for Computer Vision. Morgan & Claypool Publishers, Estados Unidos.
Kim, J. and Pavlovic, V. (2016). Discovering characteristic landmarks on ancient coins using convolutional networks. In 2016 23rd International Conference on Pattern Recognition (ICPR), pages 1595–1600, México. IEEE.
Mattioli, L. R. and Bianco, A. F. (2017). Reconhecimento de moedas brasileiras usando rede LVQ. In Proceeding Series of the Brazilian Society of Applied and Computational Mathematics, volume 5, pages 010497–1–010497–2, Gramado, RS. SBMAC.
Meneses, G. M. (2016). Contador automático de moedas utilizando redes neurais artificiais. Monografia (Bacharelado em Engenharia da Computação) – Universidade de Brası́lia.
Michelucci, U. (2019). Advanced Applied Deep Learning – Convolutional Neural Networks and Object Detection. Apress, Estados Unidos.
Moneda, L., Yonekura, D. C., and Guedes, E. B. (2020). Brazilian coin detection dataset. Disponı́vel em http://dx.doi.org/10.21227/kj4c-y809. Acesso em 24 de agosto de 2020.
Perone, C. S. (2014). Simple and effective coin segmentation using Python and OpenCV. Disponı́vel em http://blog.christianperone.com/2014/06/ simple-and-effective-coin-segmentation-using-python-and-opencv/. Acesso em 24 de agosto de 2020.
Ragab, M. E., Hassanein, A. S., Mohammad, S., and Sameer, M. (2015). A survey on Hough Transform, theory, techniques and applications. International Journal of Computer Science Issues, 12(1):139–156.
Roomi, S. M. M. and Rajee, R. B. J. (2015). Coin detection and recognition using neural networks. In 2015 International Conference on Circuits, Power and Computing Technologies [ICCPCT-2015], pages 1–6, Índia. IEEE.
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., and Chen, L.-C. (2018). MobileNetV2: Inverted residuals and linear bottlenecks. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 4510–4520, Salt Lake City, Utah, Estados Unidos. IEEE Press.
Zhang, X., Zhou, X., Lin, M., and Sun, J. (2018). Shufflenet: An extremely efficient convolutional neural network for mobile devices. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6848–6856, Salt Lake City, Estados Unidos. IEEE.
Cameron, A. C. and Windmeijer, F. A. (1997). An R-squared measure of goodness of fit for some common nonlinear regression models. Journal of Econometrics, 77(2):329–342.
Capece, N., Erra, U., and Ciliberto, A. V. (2016). Implementation of a coin recognition system for mobile devices with deep learning. In 12th International Conference on Signal-Image Technology & Internet-Based Systems (SITIS), pages 186–192, Itália. IEEE.
Capece, N., Erra, U., and Ciliberto, A. V. (2016). Implementation of a coin recognition system for mobile devices with deep learning. In 2016 12th International Conference on SignalImage Technology Internet-Based Systems (SITIS), pages 186–192, Itália. IEEE.
Chollet, F. (2017). Deep Learning with Python. Manning Publications, Estados Unidos, 1 edition. de Freitas, S. S. (2014). Determinação do valor total de moedas em imagens digitais. Mografia (Bacharelado em Sistemas de Informação) – Universidade Federal do Ceará.
Everingham, M., , Gool, L. V., Williams, C. K., Winn, J., and Zisserman, A. (2010). The Pascal Visual Object Classes (VOC) challenge. International journal of computer vision, 88(2):303–338.
Farooque, G., Sargano, A. B., Shafi, I., and Ali, W. (2016). Coin recognition with reduced feature set sift algorithm using neural network. In 2016 International Conference on Frontiers of Information Technology (FIT), pages 93–98, Paquistão. IEEE.
Freeman, I., Roese-Koerner, L., and Kummert, A. (2018). Effnet: An efficient structure for convolutional neural networks. In 2018 25th IEEE International Conference on Image Processing (ICIP), pages 6–10. IEEE.
Gupta, S., Kaur, G., Gupta, D., and Jindal, U. (2019). Brazilian coins recognition using histogram of oriented gradients features. Journal of Computational and Theoretical Nanoscience, 16(10):4170–4178.
Iandola, F. N., Han, S., Moskewicz, M. W., Ashraf, K., Dally, W. J., and Keutzer, K. (2016). Squeezenet: Alexnet-level accuracy with 50× fewer parameters and [ 0.5 mb model size. Disponı́vel em https://arxiv.org/abs/1602.07360. Acesso em 24 de agosto de 2020.
Jain, N. and Jain, N. (2012). Coin recognition using circular Hough transform. International Journal of Electronics Communication and Computer Technology, 2(3):101–104.
Jähne, B. and Haußecker, H. (2000). Computer Vision and Applications – A Guide for Students and Practitioners. Academic Press, Estados Unidos, 1 edition.
Khan, S., Rahmani, H., Shah, S. A. A., and Bennamoun, M. (2018). A Guide to Convolutional Neural Networks for Computer Vision. Morgan & Claypool Publishers, Estados Unidos.
Kim, J. and Pavlovic, V. (2016). Discovering characteristic landmarks on ancient coins using convolutional networks. In 2016 23rd International Conference on Pattern Recognition (ICPR), pages 1595–1600, México. IEEE.
Mattioli, L. R. and Bianco, A. F. (2017). Reconhecimento de moedas brasileiras usando rede LVQ. In Proceeding Series of the Brazilian Society of Applied and Computational Mathematics, volume 5, pages 010497–1–010497–2, Gramado, RS. SBMAC.
Meneses, G. M. (2016). Contador automático de moedas utilizando redes neurais artificiais. Monografia (Bacharelado em Engenharia da Computação) – Universidade de Brası́lia.
Michelucci, U. (2019). Advanced Applied Deep Learning – Convolutional Neural Networks and Object Detection. Apress, Estados Unidos.
Moneda, L., Yonekura, D. C., and Guedes, E. B. (2020). Brazilian coin detection dataset. Disponı́vel em http://dx.doi.org/10.21227/kj4c-y809. Acesso em 24 de agosto de 2020.
Perone, C. S. (2014). Simple and effective coin segmentation using Python and OpenCV. Disponı́vel em http://blog.christianperone.com/2014/06/ simple-and-effective-coin-segmentation-using-python-and-opencv/. Acesso em 24 de agosto de 2020.
Ragab, M. E., Hassanein, A. S., Mohammad, S., and Sameer, M. (2015). A survey on Hough Transform, theory, techniques and applications. International Journal of Computer Science Issues, 12(1):139–156.
Roomi, S. M. M. and Rajee, R. B. J. (2015). Coin detection and recognition using neural networks. In 2015 International Conference on Circuits, Power and Computing Technologies [ICCPCT-2015], pages 1–6, Índia. IEEE.
Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., and Chen, L.-C. (2018). MobileNetV2: Inverted residuals and linear bottlenecks. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 4510–4520, Salt Lake City, Utah, Estados Unidos. IEEE Press.
Zhang, X., Zhou, X., Lin, M., and Sun, J. (2018). Shufflenet: An extremely efficient convolutional neural network for mobile devices. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6848–6856, Salt Lake City, Estados Unidos. IEEE.
Publicado
20/10/2020
Como Citar
YONEKURA, David; GUEDES, Elloá.
Uma Abordagem Híbrida para Detecção de Moedas de Real Baseada em Algoritmos de Localização e Redes Neurais Convolucionais. In: ENCONTRO NACIONAL DE INTELIGÊNCIA ARTIFICIAL E COMPUTACIONAL (ENIAC), 17. , 2020, Evento Online.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2020
.
p. 140-151.
ISSN 2763-9061.
DOI: https://doi.org/10.5753/eniac.2020.12124.
