Abstract
The digital image matting task is an important research field of computer vision, and the method of deep image matting is a new and efficient automatic matting method. In the task of deep image matting, to solve the problem that the details of edge in the feature images of the decoder are easy to lose, the layer-skipping connection is introduced to concatenate the feature images, which have the same size in the channel dimension between the encoder and the decoder, it also realizes the fusion of shallow detailed information and deep semantic information. To get deeper semantic information and wider receptive field, the encoder uses the VGG19 network obtained by migration learning, and the decoder uses the larger convolutional kernel of 9 × 9 accordingly. At the same time, in order to solve the problems of slow speed in convergence and insufficient ability of refinement in the refined network, four convolutional layers with the residual structure are added in this network. Experimental results show that the improved network has higher accuracy and richer information of shape. The ability of generalization in this model is also stronger.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Huang, P., Zheng, Q., Liang, C.: Overview of image segmentation methods. J. Wuhan Univ. (Sci. Edn.) 66(06), 519–531 (2020)
Liang, Y., Huang, H., Cai, Z., Hao, Z., Feng, F.: Summary of natural image matting technology. Comput. Appl. Res. 38(05), 1294–1301 (2021)
Tang, J., Aksoy, Y., Oztireli, C., et al.: Learning-based sampling for natural image matting. In: Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3055–3063(2019)
Wang, J., Cohen, M.: Optimized color sampling for robust matting. In: IEEE Conference on Computer Vision & Pattern Recognition, pp. 1–8. IEEE Computer Society (2007)
Gastal, E., Oliveira, M.: Shared sampling for real-time alpha matting. Comput. Graph. Forum 29(2), 575–584 (2010)
He, K., Rhemann, C., Rother, C., et al.: A global sampling method for alpha matting. In: IEEE Conference on Computer Vision & Pattern Recognition, pp. 2049–2056. IEEE (2011)
Sun, J., Jia, J., Tang, C., et al.: Poisson matting. ACM Trans. Graph. 23(3), 315–321 (2004)
Chen, Q., Li, D., Tang, C.: KNN matting. In: IEEE Conference on Computer Vision and Pattern Recognition, vol. 35, no. 9, pp. 2175–2188. IEEE (2013)
Levin, A.: A closed form solution to natural image matting. IEEE Comput. Soc. 30(2), 228–242 (2008)
Cho, D., Tai, Y.-W., Kweon, I.: Natural image matting using deep convolutional neural networks. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) ECCV 2016. LNCS, vol. 9906, pp. 626–643. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46475-6_39
Xu, N., Price, B., Cohen, S., Huang, T.: Deep image matting. In: IEEE Conference on Computer Vision and Pattern Recognition, pp. 2970–2979 (2017)
Chen, Q., Ge, T., Xu, Y., Zhang, Z., Yang, X., Gai, K.: Semantic human matting. In: Proceedings of the 26th ACM international conference on Multimedia (2018)
Shen, X., Tao, X., Gao, H., Zhou, C., Jia, J.: Deep automatic portrait matting. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) ECCV 2016. LNCS, vol. 9905, pp. 92–107. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46448-0_6
Lutz, S., Amplianitis, K., Smolic, A.: ΑlphaGAN: generative adversarial networks for natural image matting. In: British Machine Vision Conference (2018)
Badrinarayanan, V., Kendall, A., Cipolla, R.: SegNet: a deep convolutional en-coder-decoder architecture for image segmentation. IEEE Trans. Pattern Anal. Mach. Intell. 39(12), 1 (2017)
Shelhamer, E., Long, J., Darrell, T.: Fully convolutional networks for semantic segmentation. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3431–3440 (2015)
Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. In: International Conference on Machine Learning, pp. 448–456 (2015)
He, K., Zhang, X., Ren, S., et al.: Deep residual learning for image recognition. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2016)
Xu, Z., Yang, Y.: Fast automatic portrait matting based on multitask deep learning. J. Wuhan Univ. (Eng. Edn.) 53(08), 740–745+752 (2020)
Ran, Q., Feng, J.: Automatic matting algorithm for human foreground. J. Comput. Aided Des. Graph. 32(02), 277–286 (2020)
Wang, R., Xu, S., Huang, J.: Image matting technology based on deep learning. JShanghai Univ. (Nat. Sci. Edn.) 41(05), 1–9 (2021)
Acknowledgments
This work is supported by Heilongjiang Provincial Natural Science Foundation of China (No. LH2021F034), and the Youth Innovation Talent Support Program of Harbin University of Commerce (No. 2020CX39).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Yao, G., Ma, Z. (2022). A Novel Deep Image Matting Approach Based on DIM Model. In: Hassanien, A.E., Xu, Y., Zhao, Z., Mohammed, S., Fan, Z. (eds) Business Intelligence and Information Technology. BIIT 2021. Lecture Notes on Data Engineering and Communications Technologies, vol 107. Springer, Cham. https://doi.org/10.1007/978-3-030-92632-8_40
Download citation
DOI: https://doi.org/10.1007/978-3-030-92632-8_40
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-92631-1
Online ISBN: 978-3-030-92632-8
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)