İkili Görüntülerin Sınıflandırılması için Yeni Bir Evrişimsel Sinir Ağı Mimarisi

Öz

Literatürdeki pek çok çalışma, derin öğrenme algoritmalarını kullanarak renkli görüntülerin sınıflandırılmasına odaklanırken, ikili görüntülerin sınıflandırılmasına yönelik araştırmaların sınırlı olduğu gözlemlenmektedir. İkili görüntüler için tasarlanan Evrişimli Sinir Ağ mimarileri, renkli görüntülere kıyasla genellikle daha düşük performans sergilemekle birlikte, ikili görüntülerdeki giriş verilerinin 8 bitlik renkli görüntülere göre 24 kat daha az bilgi içeriyor olması, işlem hızlarının önemli ölçüde artmasına neden olmaktadır. Bu çalışmanın amacı, yalnızca ikili görüntüler gerektiren uygulamalar —örneğin, imza tanıma, barkod okuma, QR kod tarama ve el yazısı analizi— için yüksek verimlilikle çalışan ağ mimarileri geliştirmektir. Bu hedef doğrultusunda, mevcut katmanlardan yararlanarak yeni bir Bi-CNN ağ mimarisi tasarlanmıştır. Ardından, bu mimarinin performansını artırmaya yönelik özel bir kayıp fonksiyonu geliştirilmiş ve Si-CL(İmza-Sınıflandırma) adı verilen sınıflandırma katmanı, Bi-CNN(İkili CNN)’e entegre edilerek Bi-CL-CNN olarak adlandırılan yeni bir mimari ortaya çıkmıştır. Hem Bi-CNN hem de Bi-CL-CNN, iki farklı veri kümesi üzerinde eğitilmiştir. İlk veri kümesi olan Shape-DU, bu ağları test etmek amacıyla özel olarak oluşturulmuştur. İkinci veri kümesi olan MPEG-7 ise kıyaslama amacıyla kullanılmıştır. Eğitilen ağların performansı, GoogleNet, ResNet50 ve DenseNet201 gibi daha önce eğitilmiş üç ağ ile karşılaştırılmıştır. Deneysel sonuçlar, Bi-CL-CNN ağının, doğruluk ve hesaplama hızı açısından diğer modellerden anlamlı derecede daha iyi performans gösterdiğini ortaya koymuştur. Bu bulgular, ikili görüntü veri kümelerinin işlenmesinde önerilen modellerin sağlamlık ve verimliliğini vurgulamaktadır.

Anahtar Kelimeler

• Evrişimli sinir ağı
• derin sinir ağı
• görüntü sınıflandırma
• ikili görüntü
• kayıp fonksiyonu

A Novel Convolutional Neural Network Architecture for the Classification of Binary Images

Öz

While numerous studies in the literature focus on the classification of color images using deep learning algorithms, there is a notable gap in research dedicated to the classification of binary images. Although Convolutional Neural Networks designed for binary images tend to exhibit lower performance compared to those for color images, their processing speed is significantly faster, as the input data for binary images is reduced by a factor of 24 compared with the 8-bit color images. This study aims to develop network architectures that operate with high efficiency in applications requiring only binary images, such as signature recognition, barcode reading, QR code scanning, and handwriting analysis. For this purpose, a new Bi-CNN(Binary image-CNN) network architecture was designed using existing layers. Then, a special loss function was used to improve the performance of this architecture. By integrating the classification layer called Si-CL(Signature-Classification) into Bi-CNN, a new architecture called Bi-CL-CNN emerged. Both Bi-CNN and Bi-CL-CNN were trained on two datasets. The first dataset, Shape-DU, was specifically created for testing these networks. The second dataset, MPEG-7, serves as a benchmark dataset. The performance of the trained networks is compared with three previously trained networks, namely GoogleNet, ResNet50 and DenseNet201. The empirical evaluation demonstrated that the Bi-CL-CNN network significantly outperformed the other models in both accuracy and computational speed. These findings underscore the robustness and efficiency of the proposed models in handling binary image datasets.

Anahtar Kelimeler

• Convolutional neural network
• deep neural network
• image classification
• binary image
• loss funtion

Kaynakça

1. Li, S., Song, W., Fang, L., Chen, Y., Ghamisi, P., & Benediktsson, J. A. (2019). Deep learning for hyperspectral image classification: An overview. IEEE Transactions on Geoscience and Remote Sensing, 57(9), 6690-6709. https://doi.org/10.1109/TGRS.2019.2907932
2. Das, D., Naskar, R., & Chakraborty, R. S. (2023). Image splicing detection with principal component analysis generated low-dimensional homogeneous feature set based on local binary pattern and support vector machine. Multimedia Tools and Applications, 82, 25847-25864.
3. Minaee, S., Boykov, Y., Porikli, F., Plaza, A., Kehtarnavaz, N., & Terzopoulos, D. (2022). Image segmentation using deep learning: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(7), 3523-3542.
4. Liu, L., Ouyang, W., Wang, X., Fieguth, P., Chen, J., Liu, X., & Pietikäinen, M. (2020). Deep learning for generic object detection: A survey. International Journal of Computer Vision, 128(2), 261-318.
5. Shone, N., Ngoc, T. N., Phai, V. D., & Shi, Q. (2018). A deep learning approach to network intrusion detection. IEEE Transactions on Emerging Topics in Computational Intelligence, 2(1), 41-50.
6. Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of IEEE, 86(11), 2278-2324.
7. Chollet, F. (2017). Xception: Deep learning with Depthwise separable convolutions. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1800-1807.
8. Simonyan, K., & Zisserman, A. (2015). Very deep convolutional networks for large-scale image recognition. CoRR, abs/1409, 1-14.

9. Makwe, A., & Rathore, A. S. (2021). An empirical study of neural network hyperparameters. Evolution in Computational Intelligence: Frontiers in Intelligent Computing: Theory and Applications (FICTA 2020), 1, 371-383.
10. Zhang, R., R., Zheng, Y., Mak, T. W. C., Yu, R., Wong, S. H., Lau, J. Y., & Poon, C. C. (2017). Automatic detection and classification of colorectal polyps by transferring low-level CNN features from nonmedical domain. IEEE Journal of Biomedical and Health Informatics, 21(1), 41-47. https://doi.org/10.1109/JBHI.2016.2635662
11. Ou, X., Wang, H., Liu, X., Zheng, J., Liu, Z., Tan, S., & Zhou, H. (2023). Complex scene segmentation with local to global self-attention module and feature alignment module. IEEE Access, 11, 96530-96542. https://doi.org/10.1109/ACCESS.2023.3311264
12. Jiang, J., Feng, X., Liu, F., Xu, Y., & Huang, H. (2019). Multi-spectral RGB-NIR image classification using double-channel CNN. IEEE Access, 7, 20607-20613. https://doi.org/10.1109/ACCESS.2019.2896128
13. Russ, J. C., Matey, J. R., Mallinckrodt, A. J., & McKay, S. (1994). The image processing handbook. Computers in Physics, 8(2), 177-178. https://doi.org/10.1063/1.4823282
14. Bayram, S., & Barner, K. (2023). A black-box attack on optical character recognition systems. In M. Tistarelli, S. R. Dubey, S. K. Singh, & X. Jiang (Eds.), Computer Vision and Machine Intelligence (pp. 18-34). Springer. https://doi.org/10.1007/978-981-19-7867-8_18
15. Kabakus, A. T., & Erdogmus, P. (2021). A novel handwritten Turkish letter recognition model based on convolutional neural network. Concurrency and Computation: Practice and Experience, 33(21), e6429.
16. Ojala, T., Pietikainen, M., & Maenpaa, T. (2002). Multiresolution gray-scale and rotation invariant texture classification with local binary patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, 24(7), 971-987. https://doi.org/10.1109/TPAMI.2002.1017623
17. Mujawar, S., Kiran, D., & Ramasangu, H. (2018). An efficient CNN architecture for image classification on FPGA accelerator. 2018 Second International Conference on Advances in Electronics, Computers and Communications (ICAECC), 1-4.
18. Chittajallu, S. M., Mandalaneni, N. L. D., Parasa, D., & Bano, S. (2019). Classification of binary fracture using CNN. 2019 Global Conference for Advancement in Technology (GCAT), 1-5.
19. Khandelwal, P., Khandelwal, A., & Agarwal, S. (2020). Using computer vision to enhance safety of workforce in manufacturing in a post-COVID world. arXiv. https://doi.org/10.48550/arXiv.2005.05287
20. Yang, C., Fang, L., & Wei, H. (2020). Learning contour-based mid-level representation for shape classification. IEEE Access, 8, 157587-157601. https://doi.org/10.1109/ACCESS.2020.3019800
21. Patel, V., Mujumdar, N., Balasubramanian, P., Marvaniya, S., & Mittal, A. (2019). Data augmentation using part analysis for shape classification. 2019 IEEE Winter Conference on Applications of Computer Vision (WACV), 1223-1232. https://doi.org/10.1109/WACV.2019.00135
22. Ramya, N., Om, A. S. B., & Subashini, V. (2022). Detection of pneumonia by binary image classification using hybrid neural networks. 2022 1st International Conference on Computational Science and Technology (ICCST), 1-5. https://doi.org/10.1109/ICCST55948.2022.10040322
23. Jayalakshmi, G. S., & Kumar, V. S. (2019). Convolutional neural network (CNN) based cancerous skin lesion detection system. 2019 International Conference on Computational Intelligence in Data Science (ICCIDS), 1-6. https://doi.org/10.1109/ICCIDS.2019.8862143
24. Ramanjaneyulu, K., Swamy, K. V., & Rao, C. S. (2018). Novel CBIR system using CNN architecture. 2018 3rd International Conference on Inventive Computation Technologies (ICICT), 379-383. https://doi.org/10.1109/ICICT43934.2018.9034389
25. Wang, P., Li, L., Jin, Y., & Wang, G. (2018). Detection of unwanted traffic congestion based on existing surveillance system using in freeway via a CNN-architecture trafficnet. 2018 13th IEEE Conference on Industrial Electronics and Applications (ICIEA), 1134-1139. https://doi.org/10.1109/ICIEA.2018.8397881
26. Hafeez, M. A., Munir, A., & Ullah, H. (2024). H-QNN: A hybrid quantum–classical neural network for improved binary image classification. AI, 5(3), 1462-1481.
27. Ranga, D., Prajapat, S., Akhtar, Z., Kumar, P., & Vasilakos, A. V. (2024). Hybrid quantum–classical neural networks for efficient MNIST binary image classification. Mathematics, 12(23), 1-22.
28. Korkut, Ş. G., Kocabaş, H., & Kurban, R. (2024). A comparative analysis of convolutional neural network architectures for binary image classification: A case study in skin cancer detection. Karadeniz Fen Bilimleri Dergisi, 14(4), 2008-2022.
29. Hafeez, M. A., Munir, A., & Ullah, H. (2024). H-QNN: A Hybrid Quantum–Classical Neural Network for Improved Binary Image Classification. AI, 5(3), 1462-1481.
30. Bai, Q., & Hu, X. (2023). Quantity study on a novel quantum neural network with alternately controlled gates for binary image classification. Quantum Information Processing, 22(5), 184.
31. Kiger, J., Ho, S. S., & Heydari, V. (2022, March). Malware binary image classification using convolutional neural networks. In International Conference on Cyber Warfare and Security, 17(1), 469-478. Academic Conferences International Limited. https://doi.org/10.34190/iccws.17.1.59
32. Pal, K. K., & Sudeep, K. S. (2016). Preprocessing for image classification by convolutional neural networks. 2016 IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology (RTEICT), 1778-1781. https://doi.org/10.1109/RTEICT.2016.7808140
33. Latecki, L. J., Lakamper, R., & Eckhardt, T. (2000). Shape descriptors for non-rigid shapes with a single closed contour. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1, 424-429.
34. LeCun, Y., Cortes, C., & Burges, C. (2010). MNIST handwritten digit database.
35. Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems.
36. LeCun, Y., Bottou, L., Orr, G. B., & Müller, K. R. (1998). Gradient-based learning applied to document recognition. Proceedings of IEEE.
37. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
38. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. MIT Press.
39. Szegedy, C., Vanhoucke, V., Ioffe, S., Vinyals, O., & Wu, Y. (2015). Going deeper with convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
40. Bottou, L. (2010). Large-scale machine learning with stochastic gradient descent. Proceedings of the 19th International Conference on Computational Statistics (COMPSTAT).
41. Ioffe, S., & Szegedy, C. (2015). Batch normalization: Accelerating deep network training by reducing internal covariate shift. Proceedings of the 32nd International Conference on Machine Learning.
42. Wang, Z., Zhang, L., & Wang, F. (2016). Understanding max pooling in CNNs. arXiv preprint arXiv:1509.02520.
43. Srivastava, N., Hinton, G. E., Krizhevsky, A., Sutskever, I., & Salakhutdinov, R. (2014). Dropout: A simple way to prevent neural networks from overfitting. The Journal of Machine Learning Research.
44. Ozkan, Y., & Erdogmus, P. (2024). Evaluation of classification performance of new layered convolutional neural network architecture on offline handwritten signature images. Symmetry, 16(6), 649.
45. Kingma, D. P., & Ba, J. (2015). Adam: A method for stochastic optimization. Proceedings of the 3rd International Conference on Learning Representations (ICLR).
46. Kingma, D. P., & Welling, M. (2014). Auto-Encoding variational Bayes. International Conference on Learning Representations.
47. Smith, L. N. (2017). Cyclical learning rates for training neural networks. 2017 IEEE Winter Conference on Applications of Computer Vision (WACV).
48. Keskar, N. S., Gonzalez, J. R., Dylan, Y., & Bengio, Y. (2016). On how to train a very deep neural network. Proceedings of the 5th International Conference on Learning Representations (ICLR).
49. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
50. LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of IEEE.
51. Prechelt, L. (1998). Early stopping – but when? In Neural Networks: Tricks of the Trade (pp. 55-69). Springer.
52. Pan, S. J., & Yang, Q. (2010). A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering.
53. Szegedy, C., Vanhoucke, V., Google, L. (2015). Going deeper with convolutions. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
54. Huang, G., Liu, Z., van der Maaten, L., & Weinberger, K. Q. (2017). Densely connected convolutional networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
55. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
56. Yang, E. H., Amer, H., & Jiang, Y. (2021). Compression helps deep learning in image classification. Entropy, 23(7), 881.
57. Bicego, M., & Lovato, P. (2016). A bioinformatics approach to 2D shape classification. Computer Vision and Image Understanding, 145, 59-69.
58. Govindaraj, P., & Sudhakar, M. S. (2019). A new 2D shape retrieval scheme based on phase congruency and histogram of oriented gradients. Signal, Image and Video Processing, 13, 771-778.
59. Jayasumana, S., Salzmann, M., Li, H., & Harandi, M. (2013). A framework for shape analysis via Hilbert space embedding. Proceedings of the IEEE International Conference on Computer Vision, 1249-1256.