Abstract
Underwater object tracking and recognition are challenging due to the distinctive characteristics of underwater environments. The water medium exhibits diffraction and scattering of light when it travels deep in the water. This results in unclear, occluded videos and images further causing challenges in interpretation. Tracking the object of interest from the consecutive frames in underwater scenarios often lead to occlusion of objects. Addressing these issues, an improved hybrid adaptive deep SORT-YOLOv3 (HADSYv3) detection and tracking scheme for occluded underwater objects is proposed. The neural network-based training model YOLOv3 is applied to extract and categorize the underwater object. An adaptive deep SORT algorithm with the long-short term memory (LSTM) based deep learning approach is used to determine the position of the objects in the underwater sequences. The proposed Hybrid Adaptive DeepSORT-YOLOv3 (HADSYv3) method incorporates both the YOLOv3 algorithm meant for object detection and the adaptive deep SORT algorithm for tracking applications. Though the deep SORT algorithm track objects in real time applications standalone, if it gets combined with suitable detection schemes, the overall efficiency can be improved by confirming all the possible detections. The proposed method is compared with other state of art underwater object recognition schemes and the occluded object detection for various view angles is evaluated quantitatively.
Similar content being viewed by others
References
Blanc K, Lingrand D, Precioso F (2014) Fish species recognition from video using SVM classifier. In: 3rd ACM International Workshop on Multimedia Analysis for Ecological Data, ACM, 1– 6
Carlevaris-Bianco N, Mohan A, Eustice RM (2010) September. Initial results in underwater single image dehazing. In: Oceans 2010 Mts/IEEE Seattle. IEEE, pp 1–8. https://doi.org/10.1109/OCEANS.2010.5664428
Dai J, Wang R, Zheng H, Ji G, Qiao X (2016) ZooplanktonNet: Deep convolutional network for zooplankton classification. OCEANS 2016 - Shanghai, IEEE. https://doi.org/10.1109/OCEANSAP.2016.7485680
Dhanalakshmi S, Venkatesh C (2013) Classification of ultrasound carotid artery images using texture features. International Review on Computers and Software (IRECOS) 8(4)933–940
Duan K, Keerthi SS (2005) Which is the best multiclass SVM method? Proceedings of the Sixth International Workshop on Multiple Classifier Systems, p 278– 285
Fish for Knowledge data set. Available on http://f4k.dieei.unict.it/datasets/bkgmodeling/. Accessed 1 Oct 2010 - 10 Jul 2013
Hou G-J, Luan X, Song§ D-L, Xue-Yan M (2015) Underwater Man-made Object Recognition on the basis of color and shape features. J Coastal Res Coastal Res Foundation. https://doi.org/10.2112/JCOASTRES-D-14-00249.1
Hsiao Y, Chen C, Lin S, Lin F (2014)Real-world underwater fish recognition and identification using sparse representation. Ecol Inf 23:13–21. https://doi.org/10.1016/j.ecoinf.2013.10.002
Hu W, Li X, Luo W, Zhang X, Maybank S, Zhang Z (2012) Single and multiple object tracking using log-Euclidean Riemannian subspace and block-division appearance model. IEEE Trans Pattern Anal Mach Intell 34(12):2420–2440
Kim B, Yu S (2017) Imaging sonar based real-time underwater object detection utilizing AdaBoost method. IEEE Underwater Technology (UT), Busan, pp 1–5. https://doi.org/10.1109/UT.2017.7890300
Lee D, Kim G, Kim D, Myung H, Choi H-T(2012)Vision-based object detection and tracking for autonomous navigation of underwater robots. Ocean Eng 48:59–68. ACM. https://doi.org/10.1016/j.oceaneng.2012.04.006
Lee H, Park M, Kim J (2016) Plankton classification on imbalanced large scale database via convolutional neural networks with transfer learning. 2016 IEEE International Conference on Image Processing (ICIP). https://doi.org/10.1109/ICIP.2016.7533053
Li X, Shang M, Qin H, Chen L (2015) Fast accurate fish detection and recognition of Un-derwater images with fast R-CNN. OCEANS 2015 -MTS/IEEE Washington. IEEE Press, Washington, DC, pp 1–5
Li Y, Lu H, Li J, Li X, Li Y, Serikawa S (2016) Underwater image de-scattering and classification by deep neural network. Comput Electr Eng 54:68–77, ISSN 0045-7906. https://doi.org/10.1016/j.compeleceng.2016.08.008
Mahmood A, Bennamoun M, An S, Sohel F, Boussaid F, Hovey R, Kendrick G, Fisher RB (2016) Coral Classification with hybrid feature representations. 2016 IEEE International Conference on Image Processing (ICIP). https://doi.org/10.1109/ICIP.2016.7532411
Mathias A, Samiappan D (2019) Underwater image restoration based on diffraction bounded optimization algorithm with dark channel prior. Optik 192:162925
Mathias A, Dhanalakshmi S, Kumar R, Narayanamoorthi R (2021) Underwater object detection based on bi-dimensional empirical mode decomposition and Gaussian Mixture Model approach. Ecol Inform 66:101469
Mathias A, Dhanalakshmi S, Kumar R, Narayanamoorthi R (2022) Deep neural network driven automated underwater object detection. Comput Mater Continua 70(3):5251–5267
Nikolovska (2015) AUV based flushed and buried object detection. OCEANS 2015 - Genova, Genoa, pp 1–5. https://doi.org/10.1109/OCEANS-Genova.2015.7271651
NOAA dataset available on https://videos.fisheries.noaa.gov/. Accessed 2013
Pan Y, He F, Yu H (2019) A novel enhanced collaborative autoencoder with knowledge distillation for top-N recommender systems. Neurocomputing 332:137–148
Pan Y, He F, Yu H (2020) Learning social representations with deep autoencoder for recommender system. World Wide Web 23(4):2259–2279
Quan Q, He F, Li H (2021) A multi-phase blending method with incremental intensity for training detection networks. Vis Comput 37(2):245–259
Ravanbakhsh M, Shortis M, Shafait F, Mian A, Harvey E, Seager J (2015) Automated fish detection in underwater images using shape-based level sets. Photogram Rec 30(149):46–62
Samiappan D, Chakrapani V (2016) Classification of carotid artery abnormalities in ultrasound images using an artificial neural classifier. Int Arab J Inf Technol 13(6A):756–762
Schechner YY, Karpel N (2005) Recovery of underwater visibility and structure by polarization analysis. IEEE J Oceanic Eng 30(3):570–587. https://doi.org/10.1109/JOE.2005.850871
Song B, Jeng TY, Staudt E, Roy-Chowdhury AK (2010) A stochastic graph evolution framework for robust multi-target tracking. In: European Conference on Computer Vision. Springer, Berlin, pp 605–619
Vasamsetti S, Setia S, Mittal N, Sardana HK, Babbar G (2018) Automatic underwater moving object detection using multi-feature integration framework in complex backgrounds. IET Comput Vis 12(6):770–778, The Institution of Engineering and Technology 2018
Wojke N, Bewley A, Paulus D (2017) Simple online and realtime tracking with a deep association metric. In: 2017 IEEE international conference on image processing (ICIP). IEEE, pp 3645–3649
Wright J, Yang AY, Ganesh A (2009) Robust face recognition via sparse representation. IEEE Trans Pattern Anal Mach Intell 23:210–227. https://doi.org/10.1109/TPAMI.2008.79
Yafei Zhu L, Chang J, Dai H, Zheng BZ (2016) Automatic object detection and segmentation from underwater images via saliency-based region merging. OCEANS 2016 - Shanghai, IEEE. https://doi.org/10.1109/OCEANSAP.2016.7485598
Yang H, Ping L, Hu YZ, Fu JN. Research on underwater object recognition based on YOLOv3” Microsystem Technologies. https://doi.org/10.1007/s00542-019-04694-8
Zhang L, van der Maaten L (2013) Structure preserving object tracking. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1838–1845
Zhang S, He F (2020) DRCDN: learning deep residual convolutional dehazing networks. Vis Comput 36(9):1797–1808
Zhou H, Llewellyn L, Wei L, Creighton D, Nahavandi S (2015) Marine object detection using background Modellingblob analysis. 2015 IEEE International Conference on Systems, Man, Cybernetics, IEEE. https://doi.org/10.1109/SMC.2015.86
Zu Y, Ma J, Tian J, Liu H, Yun, Zhang (2014) A gravity gradient differential ratio method for underwater object detection. IEEE Geosci Remote Sens Lett 11. https://doi.org/10.1109/LGRS.2013.2279485
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Mathias, A., Dhanalakshmi, S. & Kumar, R. Occlusion aware underwater object tracking using hybrid adaptive deep SORT -YOLOv3 approach. Multimed Tools Appl 81, 44109–44121 (2022). https://doi.org/10.1007/s11042-022-13281-5
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-13281-5