Deep Learning-based Salient Object Detection in 2D Images

Öz

Saliency detection is the process of determining how different or attention-grabbing a region within an image is compared to others. In this study, deep learning methods and image processing techniques were used for salient object detection. Using the DUTS dataset, images were augmented with horizontal flipping and random rotations at arbitrary angles. Salient object detection with convolutional neural networks (CNNs) incorporating attention mechanisms effectively focuses on both high- and low-level features. This approach has demonstrated competitive performance in detecting salient objects, and the performance of different attention mechanisms has been evaluated. Saliency object detection was performed using different deep learning models, including VGG-16, ResNet-50, and DenseNet-169.

Anahtar Kelimeler

• Salient object detection
• deep learning
• image processing

Derin Öğrenme ile 2D Görüntülerde Belirgin Nesne Tespiti

Öz

Belirginlik tespiti, bir görüntü veya bir görüntü içindeki bir bölgenin diğerlerinden ne kadar farklı veya dikkat çekici olduğunu belirleme işlemidir. Bu çalışmada, belirginlik tespiti için derin öğrenme yöntemleri ve görüntü işleme teknikleri kullanılmıştır. DUTS veri kümesi kullanılarak yapılan çalışmada; görüntülere yatay çevirme, rastgele açılarda döndürme gibi veri artırma işlemleri uygulanmıştır. Çalışmada dikkat mekanizmaları kullanılarak gerçekleştirilen Evrişimli Sinir Ağları ile belirginlik tespiti, yüksek ve düşük seviyeli özelliklere etkili bir şekilde odaklanmayı sağlamaktadır. Bu yaklaşım, belirgin nesnelerin tespiti konusunda başarılı sonuçlar elde edilmesini sağlamış ve farklı dikkat mekanizmalarının performansı değerlendirilmiştir. Çalışma kapsamında VGG-16, ResNet-50, DenseNet-169 gibi farklı derin öğrenme modellerinde belirgin nesne tespiti yapılmıştır.

Anahtar Kelimeler

• Belirgin nesne tespiti
• derin öğrenme
• görüntü işleme

Destekleyen Kurum

Bu çalışma için herhangi bir kurum ve/veya kuruluştan destek alınmamıştır.

Etik Beyan

Yazarlar çalışmanın, etik kurul izni veya herhangi bir özel izin gerektirmediğini beyan etmektedir.

Teşekkür

Bu makale Gönül Sinem Özdoğan’ın Yüksek Lisans tez çalışması kapsamında yapılmıştır.

Kaynakça

1. Wayalun, P., Chomphuwiset, P., Laopracha, N., & Wanchanthuek, P. (2012). Images enhancement of G-band chromosome using histogram equalization, OTSU thresholding, morphological dilation and flood fill techniques. 8th International Conference on Computing and Networking Technology (INC, ICCIS and ICMIC), 163-168
2. Sur, A., Sagar, S.S., Pal, R., Mitra, P., & Mukhopadhyay, J. (2009). A new image watermarking scheme using saliency based visual attention model. 2009 Annual IEEE India Conference, 1-4. https://doi.org/10.1109/INDCON.2009.5409402
3. Ren, Z., Gao, S., Chia, L. T., & Tsang, I.W.H. (2014). Region-based saliency detection and its application in object recognition. IEEE Transactions on Circuits and Systems for Video Technology, 24(5), 769-779. https://doi.org/10.1109/TCSVT.2013.2280096
4. Zhang, L., Zhang, D., Sun, J., Wei, G., & Bo, H. (2019). Salient object detection by local and global manifold regularized SVM model. Neurocomputing, 340, 42-54. https://doi.org/10.1016/j.neucom.2019.02.041
5. Liu, N., & Han, J. (2016). Dhsnet: Deep hierarchical saliency network for salient object detection. IEEE Conference on Computer Vision and Pattern Recognition (CVPR-2016), 678-686. https://doi.org/10.1109/CVPR.2016.80
6. LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521, 436-444. https://doi.org/10.1038/nature14539
7. Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017). Densely connected convolutional networks. IEEE Conference on Computer Vision and Pattern Recognition (CVPR-2017), 4700-4708. https://doi.org/10.1109/CVPR.2017.243
8. Kim, T., Kim, K., Lee, J., Cha, D., Lee, J., & Kim, D. (2023). Revisiting image pyramid structure for high resolution aalient object detection. Proceedings of the Asian Conference on Computer Vision (ACCV-2022), 257-273. https://doi.org/10.1007/978-3-031-26293-7_16

9. Zhao, T., & Wu, X. (2019). Pyramid feature attention network for saliency detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR-2019), 3080-3089. https://doi.org/10.1109/CVPR.2019.00320
10. Piao, Y., Jiang, Y., Zhang, M., Wang, J., & Lu, H. (2023). PANet: Patch-aware network for light field salient object detection. IEEE Transactions on Cybernetics, 53(1), 379-391. https://doi.org/10.1109/TCYB.2021.3095512
11. Qin, X., Zhang, Z., Huang, C., Gao, C., Dehghan, M., & Jagersand, M. (2019). Basnet: Boundary-aware salient object detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR-2019), 7479-7489. https://doi.org/10.1109/CVPR.2019.00766
12. Zhao, J., Liu, J-J., Fan, D-P., Cao, Y., Yang, J., & Cheng, M-M. (2019). EGNet: Edge guidance network for salient object detection. IEEE/CVF International Conference on Computer Vision (ICCV-2019), 8778-8787. https://doi.org/ 10.1109/ICCV.2019.00887
13. Pang, Y., Zhao, X., Zhang, L., & Lu, H. (2020). Multi-scale interactive network for salient object detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR-2020), 9410-9419. https://doi.org/10.1109/CVPR42600.2020.00943
14. Wei, J., Wang, S., & Huang, Q. (2020). F3Net: fusion, feedback and focus for salient object detection. Proceedings of the AAAI Conference on Artificial Intelligence, 34, 12321-12328. https://doi.org/10.1609/aaai.v34i07.6916
15. Liu, J. J., Hou, Q., Liu, Z. A., & Cheng, M. M. (2023). Poolnet+: Exploring the potential of pooling for salient object detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(1), 887-904. https://doi.org/10.1109/TPAMI.2021.3140168
16. Wang, Y., Wang, R., Fan, X. Wang, T., & He, X. (2023). Pixels, regions, and objects: Multiple enhancement for salient object detection. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR-2023), 10031-10040.https://doi.org/10.1109/CVPR52729.2023.00967
17. Asheghi, B., Salehpour, P., Khiavi, A. M., Hashemzadeh, M., & Monajemi, A. (2024). Dasod: Detail-aware salient object detection. Image and Vision Computing, 148, 105154. https://doi.org/10.1016/j.imavis.2024.105154
18. Wang, L., Lu, H., Wang, Y., Feng, M., Wang, D., Yin, B., & Ruan, X. (2017). Learning to detect salient objects with image-level supervision. IEEE Conference on Computer Vision and Pattern Recognition (CVPR-2017), 3796-3805. https://doi.org/10.1109/CVPR.2017.404
19. Xiong, X., Wu, Z., Tan, S., Li, W., Tang, F., Chen, Y., Li. S., Ma, J., & Li, G. (2026). Sam2-unet: Segment anything 2 makes strong encoder for natural and medical image segmentation. Visual Intelligence, 4(1), 2. https://doi.org/10.1007/s44267-025-00106-w
20. Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI-2015), 234-241. https://doi.org/10.1007/978-3-319-24574-4_28
21. Lu, Y., Chen, Y., Zhao, D., & Chen, J. (2019). Graph-FCN for image semantic segmentation. International Symposium on Neural Networks (ISNN-2019), 97-105. https://doi.org/10.1007/978-3-030-22796-8_11
22. Parhi, R., & Nowak, R. D. (2020). The role of neural network activation functions. IEEE Signal Processing Letters, 27, 1779-1783. https://doi.org/10.1109/LSP.2020.3027517
23. Üzen, H., & Fırat, H. (2024). Deep learning-based adaptive ensemble learning model for classification of monkeypox disease. Konya Journal of Engineering Sciences (Konjes), 12(4), 822-837. https://doi.org/10.36306/konjes.1471289
24. Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. International Conference on Learning Representations (ICLR-2015), 1-14. https://doi.org/10.48550/arXiv.1409.1556
25. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. IEEE Conference on Computer Vision and Pattern Recognition (CVPR-2016), 770-778. https://doi.org/10.1109/CVPR.2016.90