Advertisement

Alternative Approach to Solving Computer Vision Tasks Using Graph Structures

  • Jiajian LiEmail author
  • Mark Makarychev
  • Aleksey Popov
Chapter
Part of the Studies in Systems, Decision and Control book series (SSDC, volume 260)

Abstract

An approach to recognizing objects on images is proposed, which uses graph structures and graph algorithms. The image being processed is converted into a grid graph, which is divided into image segments using Kruskal’s algorithm and a Gaussian blur. Each resulting segment is characterized using descriptors, which are then grouped together to form the segment’s fingerprint. In the knowledge base, which is also structured as a graph, groups of object fingerprints are linked via weighted edges, which indicate the degree of contextual association. During object recognition, neighboring segments and contextual associations are used to better predict what objects are presented in the input image.

Keywords

Graphs Object recognition Computer vision Segmentation Image descriptors 

References

  1. 1.
    Liu, D., Xie, S.: Neural information processing. In: 24th International Conference, ICONIP 2017, Guangzhou, China, November 14–18, 2017, Proceedings, Part 2Google Scholar
  2. 2.
    Pechyonkin, M.: Understanding Hinton’s Capsule Networks. Part I: Intuition. [Online]. Available: https://medium.com/ai%C2%B3-theory-practice-business/understanding-hintons-capsule-networks-part-i-intuition-b4b559d1159b. Accessed: 13 June 2018
  3. 3.
    Popov, A.: An introduction to the MISD technology. In: Proceedings of the 50th Hawaii International Conference on System Sciences, HICSS50, Hawaii, 3–7 January 2017, pp. 1003–1012 (2017)Google Scholar
  4. 4.
    Ovchinnikov, V.A.: Graphs in problems of analysis and synthesis of structures of complex systems. Bauman Moscow State Technical University, Moscow (2014). [In Russian]Google Scholar
  5. 5.
    Felzenszwalb, P.F., Huttenlocher, D.P.: Efficient graph-based image segmentation. Int. J. Comput. Vis., 59(2), September 2004Google Scholar
  6. 6.
    Yamada, A., Pickering, M., Jeannin, S.: MPEG-7 visual part of experimentation model version 8.1// ISO/IEC JTC1/SC29/WG11/M6808. Pisa, Italy. January (2001) Google Scholar
  7. 7.
    Kravets, A.G., Lebedev, N., Legenchenko M.: Patents images retrieval and convolutional neural network training dataset quality improvement. In: Proceedings of the IV International research conference information technologies in Science, Management, Social sphere and medicine (ITSMSSM 2017) DEC 05–08, 2017, Tomsk, RussiaGoogle Scholar
  8. 8.
    Korobkin, D.M., Fomenkov, S.A., Kravets, A.G.: Methods for extracting the descriptions of sci-tech effects and morphological features of technical systems from patents (2019). In: 2018 9th International Conference on Information, Intelligence, Systems and Applications, IISA 2018, art. no. 8633624Google Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Bauman Moscow State Technical UniversityMoscowRussia

Personalised recommendations