Synthetic data formation for machine learning recognition of underwater objects

Synthetic data formation for machine learning recognition of underwater objects

Vyacheslav K. Abrosimov
Doctor of Technical Science, Senior Research Scientist, Federal State Budgetary Institution Main Research and Testing Interspecific Center for Advanced Weapons Ministry of Defense of the Russian Federation (GNIIMTS PV), Leading Research Scientist, 10, Chukotsky proezd, Moscow, 129327, Russia, This email address is being protected from spambots. You need JavaScript enabled to view it., ORCID: 0000-0001-6701-2389

Yulia N. Matveeva
GNIIMTS PV, Junior Research Scientist, 10, Chukotsky proezd, Moscow, 129327, Russia

Received January 23, 2023

Abstract
The main problem of machine learning for control systems of unmanned underwater vehicles is objectively very small samples of real data. The study aim was to develop an approach to the creation of synthetic data describing underwater objects for as the samples for training and validation in machine learning of control systems for autonomous unmanned underwater vehicles. The subject of the study was a variety of underwater objects because real information about the shape, size and their external images is very limited. The data augmentation method was used, which makes it possible to obtain additional data on underdetermined objects of observation from the initial data while maintaining the classification features. Eight models have been developed that imitate the influence of various factors of the aquatic environment and allow using various augmentation methods (changing the position, adding noise, glare to the image, defocusing to create fuzziness; fragmentation, etc.) to obtain an almost unlimited number of images of any objects of man-made activity immersed in underwater environment, to varying degrees similar to the reference. Examples of the use of augmentation models that take into account changes in illumination, transparency and the presence of an underwater landscape are given. Such synthetic (model) images may be the basis of a training set for machine learning to recognize and identify underwater objects. The trained model can be used as the basis of a decision support system for operators of remote-controlled unmanned underwater vehicles and as the basis for building control systems for autonomous uninhabited underwater vehicles for monitoring underwater spaces.

Key words
Machine learning, augmentation, recognition, identification modeling, remote-controlled autonomous unmanned underwater vehicle, synthetic data.

DOI
10.31776/RTCJ.11402

Bibliographic description
Abrosimov, V.K. and Matveeva, Yu.N. (2023). "Synthetic data formation for machine learning recognition of underwater objects". Robotics and Technical Cybernetics, vol. 11, no. 4, pp. 256-266, DOI: 10.31776/RTCJ.11402. (in Russian).

UDC identifier
004.8:004.93

References

  1. Vojtov, D.V. (2015), Avtonomnye neobitaemye podvodnye apparaty [Autonomous uninhabited underwater vehicles], Morkniga, Moskva, Russia, 332 p. (in Russian).
  2. Kaftannikov, I.L. and Parasich, A.V. (2016), “Problems of forming a training sample in machine learning problems”, Vestnik YUUrGU. Seriya “Komp'yuternye tekhnologii, upravlenie, radioelektronika”, vol.16, no. 3, pp. 15–24.
  3. Emel'yanov, S.O., Ivanova, A.A., SHvec, E.A. and Nikolaev, D.P. (2018), “Metody augmentacii obuchayushchih vyborok v zadachah klassifikacii izobrazhenij”, Sensory Systems, vol. 32, no. 3, pp. 236–245.
  4. ru (2019), Introduction to Image Augmentation using fastai library, [Online], available at: https://www.machinelearningmastery.ru/introduction-to-image-augmentations-using-the-fastai-library-692dfaa2da42/ (Accessed 12 May 2021).
  5. com (2021), 5 Synthetic Data Applications for the Defense Industry, [Online], available at: https://www.veteranstoday.com/2021/01/14/5-synthetic-data-applications-for-the-defense-industry/ (Accessed 3 June 2022).
  6. Akimov, A.V. (2019), “Algorithms for information processing in problems of recognition and augmentation of signals and images based on models of deforming distortions”, Abstract of PhD dissertation, 05.13.01, Voronezh, Russia, available at: http://www.science.vsu.ru/dissertations/7442/ (Accessed 15 May 2022).
  7. Osama Khalil et al. (2014), “Synthetic training in object detection”, In: Microsoft Advanced Technology Labs, Cairo, Egypt, available at: https://www.microsoft.com/en-us/research/wp-сontent/uploads/2016/02/ICIP2013_SyntheticObjectDetection_CameraReady.pdf. (Accessed 4 July 2022).
  8. Guanying Huo and Ziyin Wu (2020), “Underwater Object Classification in Sidescan Sonar Images Using Deep Transfer Learning and Semisynthetic Training Data”, IEEE Access PP(99):1-1, available at: https://www.researchgate.net/publication/339756991_Underwater_Object_Classification_in_Sidescan_Sonar_Images_Using_Deep_Transfer_Learning_and_Semisynthetic_Training_Data (Accessed 22 April 2022).
  9. Briese, Ch. and Guenther, L. (2019), “Deep Learning with Semi-Synthetic Training Images for Detection of Non-Cooperative UAVs”, Proceedings of the International Conference on Unmanned Aircraft Systems (ICUAS), available at: https://www.semanticscholar.org/paper/Deep-Learning-with-Semi-Synthetic-Training-Images-Briese-Guenther/c373f2c7104da58015d288db3bd7030959d364c2 (Accessed 13 June 2022).
  10. Golubev, P. (2019), “Application of Computer Vision in Marine Research”, Open Data Science, available at: https://habr.com/ru/company/ods/blog/454964/ (Accessed 8 May 2022).
  11. Gajer, A.V., Sheshkus, A.V. and Chernyshova, Yu.S. (2018), “On-the-fly training sample augmentation for neural network training”, Trudy ISA RAN. Specvypusk, pp. 150-157. (in Russian).
  12. Zagoruyko, S. (2015), “Learning to Compare Image Patches via Convolutional Neural Networks“, In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 4353-4361.
  13. istoriya-foto.ru (2010), Lighting and underwater photography in natural light [Online], available at: http://istoriya-foto.ru/books/item/f00/s00/z0000006/st013.shtml (Accessed 21 December 2021).
  14. RGAU-MSHA, “Introduction to the basics of underwater landscape science” [Online], available at: https://www.activestudy.info/vvedenie-v-osnovy-podvodnogo-landshaftovedeniya/ (Accessed 23 June 2022).