Abstract
While semantic segmentation networks can approximate well the shape of objects (e.g., people, chairs, tables) in images, boundary adherence is still inaccurate. We present an unsupervised image segmentation approach, named Iterative Dynamic Trees (IDT), for improved object delineation. We intend to combine IDT and semantic segmentation networks in future work towards improving object segmentation. For a given number of objects in an image graph, the IDT algorithm (i) estimates one seed per object (and background), (ii) delineates each object as one optimum-path tree, (iii) improves seed estimation and repeats steps (ii)–(iii) for a preset number of iterations or until the seed set convergence is achieved. Then, (iv) the optimum-path forest found with the lowest total cost in the loop (ii)–(iii) is selected as final segmentation. The IDT algorithm is a new method based on the Iterative Spanning Forest (ISF) framework, in which the number of superpixels is drastically reduced to the number of objects. It adds step (iv) and exploits, for the first time, dynamic arc-weight estimation in ISF for unsupervised object segmentation. We show that IDT can outperform its counterparts in two image datasets – a result that motivates its combination with semantic segmentation networks in future work.
This work was supported in part by grants from: Brazilian National Council for Scientific and Technological Development (CNPq), #313329/2020-6, #309627/2017-6, #303808/2018-7, São Paulo Research Foundation (Fapesp), #2020/09691-0, #2018/26434-0, #2014/12236-1.
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
Achanta, R., Shaji, A., Smith, K., Lucchi, A., Fua, P., Süsstrunk, S.: SLIC superpixels compared to state-of-the-art superpixel methods. IEEE Trans. Pattern Anal. Mach. Intell. 34(11), 2274–2282 (2012)
Alpert, S., Galun, M., Basri, R., Brandt, A.: Image segmentation by probabilistic bottom-up aggregation and cue integration. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, June 2007
Belém, F.C., Guimarães, S.J.F., Falcão, A.X.: Superpixel segmentation using dynamic and iterative spanning forest. IEEE Signal Process. Lett. 27, 1440–1444 (2020)
Bragantini, J., Martins, S.B., Castelo-Fernandez, C., Falcão, A.X.: Graph-based image segmentation using dynamic trees. In: Vera-Rodriguez, R., Fierrez, J., Morales, A. (eds.) CIARP 2018. LNCS, vol. 11401, pp. 470–478. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-13469-3_55
Chen, L., Papandreou, G., Kokkinos, I., Murphy, K., Yuille, A.L.: DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Trans. Pattern Anal. Mach. Intell. 40(4), 834–848 (2018)
Fahad, A., et al.: A survey of clustering algorithms for big data: taxonomy and empirical analysis. IEEE Trans Emerg. Top. Comput. 2(3), 267–279 (2014)
Falcão, A.X., Stolfi, J., de Alencar Lotufo, R.: The image foresting transform: theory, algorithms, and applications. IEEE Trans. Pattern Anal. Mach. Intell. 26(1), 19–29 (2004)
Galvão, F.L., Guimarães, S.J.F., Falcão, A.X.: Image segmentation using dense and sparse hierarchies of superpixels. Pattern Recognit. 108, 107532 (2020)
Hafiz, A.M., Bhat, G.M.: A survey on instance segmentation: state of the art. Int. J. Multimedia Inf. Retr. 1–19 (2020)
Hubert, L., Arabie, P.: Comparing partitions. J. Classif. 2(1), 193–218 (1985). https://doi.org/10.1007/BF01908075
Sofiiuk, K., Petrov, I., Barinova, O., Konushin, A.: F-BRS: rethinking backpropagating refinement for interactive segmentation. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8620–8629 (2020)
Soor, S., Challa, A., Danda, S., Daya Sagar, B., Najman, L.: Iterated watersheds, a connected variation of k-means for clustering GIS data. IEEE Trans. Emerg. Top. Comput. (2019)
Sultana, F., Sufian, A., Dutta, P.: Evolution of image segmentation using deep convolutional neural network: a survey. Knowl.-Based Syst. 201–202, 106062 (2020)
Vargas-Muñoz, J.E., Chowdhury, A.S., Alexandre, E.B., Galvão, F.L., Vechiatto Miranda, P.A., Falcão, A.X.: An iterative spanning forest framework for superpixel segmentation. IEEE Trans. Image Process. 28(7), 3477–3489 (2019)
Vinh, N.X., Epps, J., Bailey, J.: Information theoretic measures for clusterings comparison: Variants, properties, normalization and correction for chance. J. Mach. Learn. Res. 11, 2837–2854 (2010)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Aparco-Cardenas, D., de Rezende, P.J., Falcão, A.X. (2021). Object Delineation by Iterative Dynamic Trees. In: Tavares, J.M.R.S., Papa, J.P., González Hidalgo, M. (eds) Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications. CIARP 2021. Lecture Notes in Computer Science(), vol 12702. Springer, Cham. https://doi.org/10.1007/978-3-030-93420-0_13
Download citation
DOI: https://doi.org/10.1007/978-3-030-93420-0_13
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-93419-4
Online ISBN: 978-3-030-93420-0
eBook Packages: Computer ScienceComputer Science (R0)