Analysis of MobileNetV2 for fish species classification in underwater images
DOI:
https://doi.org/10.15282/mekatronika.v8i1.13187Keywords:
Fish Classification, MobileNetV2, Preprocessing methods, Image enhancementAbstract
This study investigates deep learning-based fish species classification in underwater images using the MobileNetV2 architecture combined with sophisticated data augmentation techniques, such as rotation and scaling. Robust preprocessing methods, including noise reduction, contrast enhancement, and normalisation, were employed as preprocessing steps. The system facilitates seamless dataset loading and classification procedures via a user-friendly GUI, enabling effective underwater image enhancement and accurate fish classification. Evaluation metrics such as accuracy, precision, recall, F1-score, and k-fold cross-validation were used to assess the project's performance, offering insights into the model's capabilities and generalisability across various underwater conditions. In conclusion, MobileNetV2, a novel underwater image enhancement and fish classification model, has the potential to revolutionise aquaculture. However, it faces limitations, such as a limited preprocessing pipeline owing to underwater conditions and the need for adapted data augmentation techniques. Solutions include developing specific techniques for different underwater environments, optimising annotation processes, and using transfer learning methods. Challenges in dataset availability and complexity include inadequate datasets, computational complexity, and complex annotation and classification difficulties. Alternatives include developing preprocessing methods specifically for underwater environments, using model compression techniques, improving the annotations, customising larger datasets, and leveraging transfer learning.
References
[1] Sandler M, Howard A, Zhu M, Zhmoginov A, Chen LC. MobileNetV2: Inverted residuals and linear bottlenecks. In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition; 2018: 4510-4520. https://doi.org/10.1109/cvpr.2018.00474
[2] Xu S, Zhang M, Song W, Mei H, He Q, Liotta A. A systematic review and analysis of deep learning-based underwater object detection. Neurocomputing. 2023;527;204–232. https://doi.org/10.1016/j.neucom.2023.01.056
[3] Tang Y, Iwaguchi T, Kawasaki H, Sagawa R, Furukawa R. AutoEnhancer: Transformer on U-net architecture search for underwater image enhancement. In: Wang L, Gall J, Chin TJ, Sato I, Chellappa R, editors. Computer Vision – ACCV 2022. Lecture Notes in Computer Science. Vol. 13846. Cham: Springer; 2023: 120–137. https://doi.org/10.1007/978-3-031-263132_8
[4] Haider A, Arsalan M, Nam SH, Sultan H, Park KR. Computer-aided fish assessment in an underwater marine environment using parallel and Progressive Spatial Information Fusion. Journal of King Saud University - Computer and Information Sciences, 2023;35(3):211–226. https://doi.org/10.1016/j.jksuci.2023.02.016
[5] Zheng S, Lin X, Zhang W, He B, Jia S, Wang P, Jia F. MDCC-Net: multiscale double-channel convolution U-Net framework for colorectal tumor segmentation. Computers in Biology and Medicine, 2021;130:104183. https://doi.org/10.1016/j.compbiomed.2020.104183
[6] Teli SG, Shelke RJ. Review of Underwater Image Enhancement using CNN and U-net. Journal of Artificial Intelligence and Imaging. 2023:1(1);5–10. https://doi.org/10.48001/joaii.2023.115-10
[7] Mohd Azmi KZ, Abdul Ghani AS, Md Yusof Z, Mohammad-Noor N, Ismail HN, Abu MY. 2024. Underwater image contrast enhancement through an intensity-randomised approach incorporating a swarm intelligence technique with unsupervised dual-step fusion. International Journal of Image and Data Fusion. 2024;15(3):310-344. https://doi.org/10.1080/19479832.2023.2289490
[8] Liu Y, An D, Ren Y, Zhao J, Zhang C, Cheng J, Liu J, Wei Y. DP-FishNet:Dual-path pyramid vision transformer-based Underwater Fish Detection Network. Expert Systems with Applications. 2024;238:122018. https://doi.org/10.1016/j.eswa.2023.122018
[9] Huang H, Lin L, Tong R, Hu H, Zhang Q, Iwamoto Y, et al. UNet 3+: A full-scale connected unet for medical image segmentation. In: ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing; 2020: 1055-1059. https://doi.org/10.1109/icassp40776.2020.9053405
[10] Chuang YC., Hu Z Han B Wang P, Zhao Y, Wu H. . Intelligent Measurement of Morphological Characteristics of Fish Using Improved UNet. Electronics. 2021;10(12):1426. https://doi.org/10.3390/electronics10121426
[11] Sfakianakis C, Simantiris G, Tziritas G. Gudu: Geometrically constrained ultrasound data augmentation in U-Net for echocardiography semantic segmentation. Biomedical Signal Processing and Control 2023; 82:104557. https://doi.org/10.1016/j.bspc.2022.104557
[12] Kamarul Baharin MAS, Abdul Ghani AS, Syamsul Amri SQ, Mohammad-Noor N, Ismail HN. Enhancement of Low-Quality Diatom Images using Integrated Automatic Background Removal (IABR) ethod from Digital Microscopic Image. In: 2023 IEEE 13th Symposium on Computer Applications & Industrial Electronics; 2023: 256-261 https://doi.org/10.1109/ISCAIE57739.2023.10165175
[13] Zhou J, et al. UGIF-Net: An Efficient Fully Guided Information Flow Network for Underwater Image Enhancement. IEEE Transactions on Geoscience and Remote Sensing. 2023;61:1-17. Article 4206117. https://doi.org/10.1109/TGRS.2023.3293912
[14] Ben Tamou A, Benzinou A, Nasreddine K 2022. Targeted Data Augmentation and Hierarchical Classification with Deep Learning for Fish Species Identification in Underwater Images. Journal of Imaging. 2022;8(8):214. https://doi.org/10.3390/jimaging8080214
[15] Wang X, Mao L, Wang J. Underwater Fish Image Enhancement method based on color correction. Highlights in Science, Engineering and Technology. 2022;1:415–425. https://doi.org/10.54097/hset.v1i.498
[16] Mohd Azmi KZ, Abdul Ghani AS, Md Yusof Z. Red and blue channels correction based on green channel and median-based dual-intensity images fusion for turbid underwater image quality enhancement. In: Md Zain Z, et al., editors. Proceedings of the 11th National Technical Seminar on Unmanned System Technology 2019. Lecture Notes in Electrical Engineering. Vol. 666. Singapore: Springer; 2021: 73–86. https://doi.org/10.1007/978-981-15-5281-6_6
[17] Hambarde P, Murala S, Dhall A, UW-GAN: Single-Image Depth Estimation and Image Enhancement for Underwater Images. IEEE Transactions on Instrumentation and Measurement. 2021;70:1–12. Article 5018412. https://doi.org/10.1109/TIM.2021.3120130
[18] Fitrianah D, Suryaningrum K M, Sagala NT, Ayumi V, Lim SM. Fine-tuned mobilenetv2 and VGG16 algorithm for Fish Image Classification. In: 2022 International Conference on Informatics, Multimedia, Cyber and Information System; 2022: 384–389. https://doi.org/10.1109/ICIMCIS56303.2022.10017901
[19] Liu X, Jia Z, Hou X, Fu M, Ma L, Sun Q. Real-time marine animal images classification by embedded system based on Mobilenet and Transfer Learning. In: OCEANS 2019 - Marseille; 2019: 1–5. https://doi.org/10.1109/OCEANSE.2019.8867190
[20] Kamarul Baharin MAS, Abdul Ghani AS, Mohammad-Noor N, Ismail HN, Syamsul Amri SQ. Automatic phytoplankton image smoothing through integrated dual image histogram specification and enhanced background removal method. The Imaging Science Journal. 2021;69(1-4), 65-91. https://doi.org/10.1080/13682199.2022.2149067
Downloads
Published
Issue
Section
License
Copyright (c) 2026 The Author(s)
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
