# Underwater Image Classification Algorithm Based on Convolutional Neural Network and Optimized Extreme Learning Machine

^{1}

^{2}

^{3}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Related Work

#### 2.1. Extreme Learning Machine

#### 2.2. Flow Direction Algorithm

## 3. Underwater Image Classification Model Based on CNN and Optimized ELM

#### 3.1. Chaos Initialization

#### 3.2. Flow Direction Algorithm Based on Chaos Initialization and Multi Population Strategy

#### 3.3. Fuzzy Logic for FDA

#### 3.4. Search Agent Strategy

#### 3.5. Flow of FCMFDA-ELM Underwater Image Classification Algorithm

## 4. Experimental Results and Analysis

#### 4.1. Datasets

#### 4.2. Performance Evaluation Indicators

#### 4.3. Experimental Parameter Setting

#### 4.4. Discussion on Experimental Results

#### 4.4.1. Experimental Analysis on Fish4Knowledge Dataset

#### 4.4.2. Experimental Analysis on URPC Dataset

^{−2}, which is greater than 0.01, but still conforms to p < 0.05. It can reflect the differentiation, which is enough to prove the excellence of FCMFDA-ELM.

## 5. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- Zhou, J.; Yang, T.; Zhang, W. Underwater vision enhancement technologies: A comprehensive review, challenges, and recent trends. Appl. Intell.
**2022**, 1–28. Available online: https://link.springer.com/article/10.1007/s10489-022-03767-y (accessed on 1 October 2022). [CrossRef] - Blake, R.E. Digital image processing of underwater images. In Proceedings of the 2006 IEEE US/EU Baltic International Symposium on Integrated Ocean Observation Syst. for Managing Global & Regional Ecosys, Marine Resch, Klaipeda, Lithuania, 23–26 May 2006. [Google Scholar]
- Han, F.; Yao, J.; Zhu, H.; Wang, C. Underwater Image Processing and Object Detection Based on Deep CNN Method. J. Sens.
**2020**, 2020, 1–20. [Google Scholar] [CrossRef] - Burguera, A. Segmentation through patch classification: A neural network approach to detect Posidonia oceanica in underwater images. Ecol. Inform.
**2020**, 56, 101053. [Google Scholar] [CrossRef] - Qiao, X.; Bao, J.; Zeng, L.; Zou, J.; Li, D. An automatic active contour method for sea cucumber segmentation in natural underwater environments. Comput. Electron. Agric.
**2017**, 135, 134–142. [Google Scholar] [CrossRef] - Spampinato, C.; Giordano, D.; Di Salvo, R.; Chen-Burger, Y.H.J.; Fisher, R.B.; Nadarajan, G. Automatic fish classification for underwater species behavior understanding. In Proceedings of the ARTEMIS’10—Proceedings of the 1st ACM Workshop on Analysis and Retrieval of Tracked Events and Motion in Imagery Streams, Co-Located with ACM Multimedia, Firenze, Italy, 29 October 2010; pp. 45–50. [Google Scholar]
- Huang, P.X.; Boom, B.J.; Fisher, R.B. Hierarchical classification with reject option for live fish recognition. Mach. Vis. Appl.
**2014**, 26, 89–102. [Google Scholar] [CrossRef] [Green Version] - Wang, C.-N.; Yang, F.-C.; Nguyen, V.T.T.; Vo, N.T.M. CFD Analysis and Optimum Design for a Centrifugal Pump Using an Effectively Artificial Intelligent Algorithm. Micromachines
**2022**, 13, 1208. [Google Scholar] [CrossRef] [PubMed] - Le Chau, N.; Dao, T.-P.; Nguyen, V.T.T. An Efficient Hybrid Approach of Finite Element Method, Artificial Neural Network-Based Multiobjective Genetic Algorithm for Computational Optimization of a Linear Compliant Mechanism of Nanoindentation Tester. Math. Probl. Eng.
**2018**, 2018, 1–19. [Google Scholar] [CrossRef] - Krizhevsky, A.; Sutskever, I.; Hinton, G.E. Imagenet classification with deep convolutional neural networks. Commun. ACM
**2017**, 60, 84–90. [Google Scholar] [CrossRef] [Green Version] - Ríos-Muñoz, G.R.; Fernández-Avilés, F.; Arenal, Á. Convolutional Neural Networks for Mechanistic Driver Detection in Atrial Fibrillation. Int. J. Mol. Sci.
**2022**, 23, 4216. [Google Scholar] [CrossRef] - Villon, S.; Mouillot, D.; Chaumont, M.; Darling, E.S.; Subsol, G.; Claverie, T.; Villéger, S. A Deep learning method for accurate and fast identification of coral reef fishes in un-derwater images. Ecol. Inform.
**2018**, 48, 238–244. [Google Scholar] [CrossRef] - Salman, A.; Jalal, A.; Shafait, F.; Mian, A.; Shortis, M.; Seager, J.; Harvey, E. Fish species classification in unconstrained underwater environments based on deep learning. Limnol. Oceanogr. Methods
**2016**, 14, 570–585. [Google Scholar] [CrossRef] [Green Version] - Qin, H.; Li, X.; Liang, J.; Peng, Y.; Zhang, C. DeepFish: Accurate underwater live fish recognition with a deep architecture. Neurocomputing
**2016**, 187, 49–58. [Google Scholar] [CrossRef] - Labao, A.B.; Naval, P.C., Jr. Cascaded deep network systems with linked ensemble components for underwater fish detection in the wild. Ecol. Inform.
**2019**, 52, 103–121. [Google Scholar] [CrossRef] - Hu, J.; Li, D.; Duan, Q.; Han, Y.; Chen, G.; Si, X. Fish species classification by color, texture and multi-class support vector machine using computer vision. Comput. Electron. Agric.
**2012**, 88, 133–140. [Google Scholar] [CrossRef] - Guo, X.; Zhao, X.; Liu, Y.; Li, D. Underwater sea cucumber identification via deep residual networks. Inf. Process. Agric.
**2019**, 6, 307–315. [Google Scholar] [CrossRef] - Prasetyo, E.; Suciati, N.; Fatichah, C. Multi-level residual network VGGNet for fish species classification. J. King Saud Univ.-Comput. Inf. Sci.
**2022**, 34, 5286–5295. [Google Scholar] [CrossRef] - Gómez-Ríos, A.; Tabik, S.; Luengo, J.; Shihavuddin, A.; Herrera, F. Coral species identification with texture or structure images using a two-level classifier based on Convolutional Neural Networks. Knowl.-Based Syst.
**2019**, 184, 104891. [Google Scholar] [CrossRef] [Green Version] - Kumar, K.A.; Prasad, A.Y.; Metan, J. A hybrid deep CNN-Cov-19-Res-Net Transfer learning architype for an enhanced Brain tumor Detec-tion and Classification scheme in medical image processing. Biomed. Signal Process. Control
**2022**, 76, 103631. [Google Scholar] [CrossRef] - Ji, Y.; Zhang, H.; Wu, Q.J. Salient object detection via multi-scale attention CNN. Neurocomputing
**2018**, 322, 130–140. [Google Scholar] [CrossRef] - Wang, X.; Jiao, J.; Yin, J.; Zhao, W.; Han, X.; Sun, B. Underwater sonar image classification using adaptive weights convolutional neural network. Appl. Acoust.
**2018**, 146, 145–154. [Google Scholar] [CrossRef] - Zhang, K.; Wang, W.; Lv, Z.; Fan, Y.; Song, Y. Computer vision detection of foreign objects in coal processing using attention CNN. Eng. Appl. Artif. Intell.
**2021**, 102, 104242. [Google Scholar] [CrossRef] - Gómez-Ríos, A.; Tabik, S.; Luengo, J.; Shihavuddin, A.; Krawczyk, B.; Herrera, F. Towards highly accurate coral texture images classification using deep convolutional neural networks and data augmentation. Expert Syst. Appl.
**2018**, 118, 315–328. [Google Scholar] [CrossRef] [Green Version] - Dutta, M.K.; Sengar, N.; Kamble, N.; Banerjee, K.; Minhas, N.; Sarkar, B. Image processing based technique for classification of fish quality after cypermethrine exposure. LWT
**2016**, 68, 408–417. [Google Scholar] [CrossRef] - Alshdaifat, N.F.F.; Talib, A.Z.; Osman, M.A. Improved deep learning framework for fish segmentation in underwater videos. Ecol. Inform.
**2020**, 59, 101121. [Google Scholar] [CrossRef] - Hu, G.; Wang, K.; Peng, Y.; Qiu, M.; Shi, J.; Liu, L. Deep Learning Methods for Underwater Target Feature Extraction and Recognition. Comput. Intell. Neurosci.
**2018**, 2018, 1214301. [Google Scholar] [CrossRef] - Khellal, A.; Ma, H.; Fei, Q. Convolutional neural network based on extreme learning machine for maritime ships recognition in in-frared images. Sensors
**2018**, 18, 1490. [Google Scholar] [CrossRef] [Green Version] - Li, J.; Shi, W.; Yang, D. Clothing Image Classification with a Dragonfly Algorithm Optimised Online Sequential Extreme Learning Machine. Fibres Text. East. Eur.
**2021**, 29, 91–96. [Google Scholar] [CrossRef] - Li, J.; Shi, W.; Yang, D. Color difference classification of dyed fabrics via a kernel extreme learning machine based on an im-proved grasshopper optimization algorithm. Color Res. Appl.
**2021**, 46, 388–401. [Google Scholar] [CrossRef] - Li, J.; Shi, W.; Yang, D. Fabric wrinkle evaluation model with regularized extreme learning machine based on improved Harris Hawks optimization. J. Text. Inst.
**2021**, 113, 199–211. [Google Scholar] [CrossRef] - Huang, G.-B.; Zhu, Q.-Y.; Siew, C.-K. Extreme learning machine: Theory and applications. Neurocomputing
**2006**, 70, 489–501. [Google Scholar] [CrossRef] - Huang, G.-B.; Chen, L. Convex incremental extreme learning machine. Neurocomputing
**2007**, 70, 3056–3062. [Google Scholar] [CrossRef] - Liang, N.Y.; Huang, G.B.; Saratchandran, P.; Sundararajan, N. A fast and accurate online sequential learning algorithm for feedforward networks. IEEE Trans. Neural Netw.
**2006**, 17, 1411–1423. [Google Scholar] [CrossRef] [PubMed] - Ganesan, A.; Santhanam, S.M. A novel feature descriptor based coral image classification using extreme learning machine with ameliorated chimp optimization algorithm. Ecol. Inform.
**2022**, 68, 101527. [Google Scholar] [CrossRef] - Huang, G.; Liu, Z.; Van Der Maaten, L.; Weinberger, K.Q. Densely connected convolutional networks. In Proceedings of the 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017, Honolulu, HI, USA, 21–26 July 2017; pp. 2261–2269. [Google Scholar]
- Lin, W.-H.; Zhong, J.-X.; Liu, S.; Li, T.; Li, G. ROIMIX: Proposal-Fusion Among Multiple Images for Underwater Object Detection. In Proceedings of the ICASSP 2020–2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Barcelona, Spain, 4–8 May 2020; pp. 2588–2592. [Google Scholar] [CrossRef] [Green Version]
- Karami, H.; Anaraki, M.V.; Farzin, S.; Mirjalili, S. Flow Direction Algorithm (FDA): A Novel Optimization Approach for Solving Optimization Problems. Comput. Ind. Eng.
**2021**, 156, 107224. [Google Scholar] [CrossRef] - Nguyen, T.V.T.; Huynh, N.-T.; Vu, N.-C.; Kieu, V.N.D.; Huang, S.-C. Optimizing compliant gripper mechanism design by employing an effective bi-algorithm: Fuzzy logic and ANFIS. Microsyst. Technol.
**2021**, 27, 3389–3412. [Google Scholar] [CrossRef]

**Figure 4.**Flow chart of underwater image classification algorithm based on convolution network and ELM optimized by FCMFDA.

**Figure 5.**Structure diagram of underwater image classification network based on convolution neural network and FCMFDA-ELM.

**Figure 8.**Confusion matrix (Fish4Knowledge) of each algorithm, including DenseNet201 (

**A**), FCMFDA-ELM (

**B**), FDA-ELM (

**C**), STOA-ELM (

**D**), WOA-ELM (

**E**), MFO-ELM (

**F**), and ELM (

**G**).

**Figure 13.**Confusion matrix (URPC) of each algorithm, including DenseNet201 (

**A**), FCMFDA-ELM (

**B**), FDA-ELM (

**C**), STOA-ELM (

**D**), WOA-ELM (

**E**), MFO-ELM (

**F**), and ELM (

**G**).

Name | Map |
---|---|

Chebyshev map | ${x}_{i+1}=cos\left(ico{s}^{-1}\left({x}_{i}\right)\right)$ |

Iterative map | ${x}_{i+1}=sin\left(\frac{a\pi}{{x}_{i}}\right),a\in \left(0,1\right)$ |

Logistic map | ${x}_{i+1}=a{x}_{i}\left(1-{x}_{i}\right)$ |

Sine map | ${x}_{i+1}=\frac{a}{4}sin\left(\pi {x}_{i}\right),a\in \left(0,4\right]$ |

Singer map | ${x}_{i+1}=\mu \left(7.86{x}_{i}-23.31{x}_{i}{}^{2}+28.75{x}_{i}{}^{3}-13.302875{x}_{i}{}^{4}\right)$ |

Sinusoidal map | ${x}_{i+1}=\mathrm{a}{x}_{i}{}^{2}sin\left(\pi {x}_{i}\right)$ |

Tent map | ${x}_{i+1}=\{\begin{array}{c}\frac{{x}_{i}}{0.7},{x}_{i}0.7\\ \frac{10}{3}\left(1-{x}_{i}\right),{x}_{i}\ge 0.7\end{array}$ |

Type | Function | Range | Dim | MinValue |
---|---|---|---|---|

Unimodal | $f1={\Sigma}_{i=1}^{n}{x}_{i}^{2}$ | [−100, 100] | 30 | 0 |

$f3={\Sigma}_{i=1}^{n}{\left({\Sigma}_{j-1}^{i}{x}_{j}\right)}^{2}$ | [−100, 100] | 30 | 0 | |

$f5={\Sigma}_{i=1}^{n}\left[100{\left({x}_{i+1}-{x}_{i}^{2}\right)}^{2}+{\left({x}_{i}-1\right)}^{2}\right]$ | [−30, 30] | 30 | 0 | |

Multimodal | $f9={\Sigma}_{i=1}^{n}\left[{x}_{i}^{2}-10cos\left(2\pi {x}_{i}\right)+10\right]$ | [−5.12, 5.12] | 30 | 0 |

$f11=\frac{1}{4000}{\Sigma}_{i=1}^{n}{x}_{i}^{2}-{\Pi}_{i=1}^{n}cos\left(\frac{{x}_{i}}{\sqrt{i}}\right)+1$ | [−600, 600] | 30 | 0 | |

$\begin{array}{cc}\hfill f13& =0.1\{{sin}^{2}\left(3\pi {x}_{1}\right)\hfill \\ & +{\Sigma}_{i=1}^{n}{\left({x}_{i}-1\right)}^{2}[1\hfill \\ & +{sin}^{2}\left(3\pi {x}_{1}+1\right)]\hfill \\ & +{\left({x}_{n}-1\right)}^{2}\left[1+{sin}^{2}\left(2\pi {x}_{n}\right)\right]\}\hfill \\ & +{\Sigma}_{i=1}^{n}u\left({x}_{i},5,100,4\right)\hfill \end{array}$ | [−50, 50] | 30 | 0 | |

Fixed-dimension multimodal | $f16=4{x}_{1}^{2}-2.1{x}_{1}^{4}+\frac{1}{3}{x}_{1}^{6}+{x}_{1}{x}_{2}-4{x}_{2}^{2}+4{x}_{2}^{4}$ | [−5, 5] | 2 | −1.0316 |

$\begin{array}{cc}\hfill f18=[1+& {({x}_{1}+{x}_{2}+1)}^{2}(19-4{x}_{1}+3{x}_{1}^{2}\hfill \\ & -14{x}_{2}+6{x}_{1}{x}_{2}+3{x}_{2}^{2}\left)\right]\hfill \\ & \ast [30+{\left(2{x}_{1}-3{x}_{2}\right)}^{2}(18\hfill \\ & -32{x}_{1}+12{x}_{1}^{2}+48{x}_{2}\hfill \\ & -36{x}_{1}{x}_{2}+27{x}_{2}^{2}\left)\right]\hfill \end{array}$ | [−2, 2] | 2 | 3 | |

$f20=-{\Sigma}_{i=1}^{4}{c}_{i}\mathrm{exp}(-{\Sigma}_{j=1}^{6}{a}_{ij}{\left({x}_{j}-{p}_{ij}\right)}^{2})$ | [0, 1] | 6 | −3.32 |

Logistic Map | Chebyshev Map | Iterative Map | Sine Map | Singer Map | Sinusoidal Map | Tent Map | |
---|---|---|---|---|---|---|---|

F1 | 0.00 × 10^{0} | 1.89 × 10^{−157} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 1.76 × 10^{−127} | 0.00 × 10^{0} |

F3 | 0.00 × 10^{0} | 2.80 × 10^{−168} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 4.61 × 10^{−92} | 4.51 × 10^{−97} |

F5 | 1.40 × 10^{1} | 1.61 × 10^{1} | 2.01 × 10^{1} | 5.91 × 10^{−7} | 2.23 × 10^{1} | 2.25 × 10^{1} | 1.81 × 10^{1} |

F9 | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} |

F11 | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} | 0.00 × 10^{0} |

F13 | 8.50 × 10^{−22} | 1.10 × 10^{−2} | 9.74 × 10^{−2} | 1.10 × 10^{−2} | 1.10 × 10^{−2} | 2.10 × 10^{−2} | 5.48 × 10^{−2} |

F16 | −1.03 × 10^{0} | −1.03 × 10^{0} | −1.03 × 10^{0} | −1.03 × 10^{0} | −1.03 × 10^{0} | −1.03 × 10^{0} | −1.03 × 10^{0} |

F18 | 3.00 × 10^{0} | 3.00 × 10^{0} | 3.00 × 10^{0} | 3.00 × 10^{0} | 3.00 × 10^{0} | 3.00 × 10^{0} | 3.00 × 10^{0} |

F20 | −3.32 × 10^{0} | −3.32 × 10^{0} | −3.32 × 10^{0} | −3.32 × 10^{0} | −3.32 × 10^{0} | −3.32 × 10^{0} | −3.32 × 10^{0} |

Logistic Map | Chebyshev Map | Iterative Map | Sine Map | Singer Map | Sinusoidal Map | Tent Map | |
---|---|---|---|---|---|---|---|

Score ranking | 25 | 19 | 16 | 23 | 18 | 11 | 16 |

Main loop of multi-population |

For i in number of multi-populationsIf (MaxFitness (overall) < MaxFitness in population(i))bestFlow ( overall) = bestFlow in population(i)EndIfEndFor |

No | Rules |
---|---|

1 | NFV is Low, ρ is Low → Δρ is PO |

2 | NFV is Low, ρ is Med → Δρ is PO |

3 | NFV is Low, ρ is High → Δρ is ZE |

4 | NFV is Med, ρ is Low → Δρ is PO |

5 | NFV is Med, ρ is Med → Δρ is ZE |

6 | NFV is Med, ρ is High → Δρ is NE |

7 | NFV is High, ρ is Low → Δρ is ZE |

8 | NFV is High, ρ is Med → Δρ is ZE |

9 | NFV is High, ρ is High → Δρ is NE |

Parameter | FCMFDA-ELM | FDA-ELM | STOA-ELM | WOA-ELM | MFO-ELM | ELM |
---|---|---|---|---|---|---|

Population size | 20 | 20 | 20 | 20 | 20 | 20 |

Maximum number of iterations | 150 | 150 | 150 | 150 | 150 | 150 |

Additional population | 3 | -- | -- | -- | -- | -- |

Maximum number of input nodes (Fish4/URPC) | 587/480 | 587/480 | 587/480 | 587/480 | 587/480 | 587/480 |

Number of hidden layer nodes (Fish4/URPC) | 100 | 100 | 100 | 100 | 100 | 100 |

Number of tributaries β | 1 | 1 | -- | -- | -- | -- |

Particle search range | [−1, 1] | [−1, 1] | [−1, 1] | [−1, 1] | [−1, 1] | [−1, 1] |

Logarithmic helix shape constant b | -- | -- | -- | -- | 1 | -- |

Method | Norm | |||||
---|---|---|---|---|---|---|

n | m | Precision | Recall | Accuracy | F1 | |

DenseNet201 | -- | -- | 0.9018 ± 0.0031 | 0.8600 ± 0.0012 | 0.9000 ± 0.0183 | 0.8263 ± 0.0312 |

FCMFDA-ELM | 337 | 52 | 0.9951 ± 0.0016 | 0.9948 ± 0.0016 | 0.9990 ± 0.0001 | 0.9948 ± 0.0011 |

FDA-ELM | 229 | 36 | 0.9604 ± 0.0353 | 0.9610 ± 0.0348 | 0.9606 ± 0.0346 | 0.9593 ± 0.0362 |

STOA-ELM | 229 | 37 | 0.9568 ± 0.0369 | 0.9567 ± 0.0388 | 0.9567 ± 0.0381 | 0.9950 ± 0.0395 |

WOA-ELM | 380 | 60 | 0.9800 ± 0.0168 | 0.9802 ± 0.0163 | 0.9801 ± 0.0152 | 0.9796 ± 0.0164 |

MFO-ELM | 303 | 35 | 0.9799 ± 0.0172 | 0.9796 ± 0.0172 | 0.9797 ± 0.0158 | 0.9790 ± 0.0172 |

ELM | 322 | 70 | 0.9348 ± 0.474 | 0.9355 ± 0.0475 | 0.9356 ± 0.0465 | 0.9300 ± 0.0478 |

Ours | Others | p-Value | |||
---|---|---|---|---|---|

Precision | Recall | Accuracy | F1 | ||

FCMFDA-ELM | FDA-ELM | 4.363 × 10^{−4} | 7.011 × 10^{−4} | 3.746 × 10^{−4} | 5.221 × 10^{−4} |

STOA-ELM | 4.498 × 10^{−6} | 4.416 × 10^{−6} | 1.832 × 10^{−6} | 3.059 × 10^{−6} | |

WOA-ELM | 1.382 × 10^{−3} | 8.472 × 10^{−3} | 1.712 × 10^{−2} | 7.732 × 10^{−3} | |

MFO-ELM | 1.571 × 10^{−4} | 9.785 × 10^{−5} | 2.464 × 10^{−5} | 1.723 × 10^{−4} | |

ELM | 7.111 × 10^{−8} | 2.724 × 10^{−8} | 7.731 × 10^{−9} | 4.371 × 10^{−8} |

Method | Norm | |||||
---|---|---|---|---|---|---|

n | m | Precision | Recall | Accuracy | F1 | |

DenseNet201 | -- | -- | 0.8718 ± 0.0114 | 0.8924 ± 0.0086 | 0.8884 ± 0.0024 | 0.8770 ± 0.0035 |

FCMFDA-ELM | 258 | 63 | 0.9682 ± 0.0050 | 0.9654 ± 0.0061 | 0.9852 ± 0.0024 | 0.9667 ± 0.0055 |

FDA-ELM | 210 | 35 | 0.9309 ± 0.0156 | 0.9246 ± 0.0127 | 0.9428 ± 0.0076 | 0.9274 ± 0.0141 |

STOA-ELM | 202 | 33 | 0.9505 ± 0.0090 | 0.9531 ± 0.0147 | 0.9556 ± 0.0045 | 0.9522 ± 0.0070 |

WOA-ELM | 254 | 35 | 0.9419 ± 0.0080 | 0.9340 ± 0.0072 | 0.9440 ± 0.0062 | 0.9376 ± 0.0073 |

MFO-ELM | 264 | 31 | 0.9366 ± 0.0240 | 0.9305 ± 0.0318 | 0.9410 ± 0.0244 | 0.9325 ± 0.0302 |

ELM | 240 | 62 | 0.9316 ± 0.0199 | 0.9242 ± 0.0203 | 0.9330 ± 0.0211 | 0.9275 ± 0.0200 |

Ours | Others | p-Value | |||
---|---|---|---|---|---|

Precision | Recall | Accuracy | F1 | ||

FCMFDA-ELM | FDA-ELM | 2.714 × 10^{−6} | 6.346 × 10^{−8} | 1.251 × 10^{−8} | 3.782 × 10^{−7} |

STOA-ELM | 5.967 × 10^{−5} | 4.571 × 10^{−2} | 9.835 × 10^{−7} | 1.234 × 10^{−4} | |

WOA-ELM | 8.351 × 10^{−8} | 3.331 × 10^{−9} | 3.221 × 10^{−9} | 7.713 × 10^{−9} | |

MFO-ELM | 1.316 × 10^{−3} | 6.263 × 10^{−3} | 3.277 × 10^{−3} | 4.558 × 10^{−3} | |

ELM | 4.623 × 10^{−5} | 2.251 × 10^{−5} | 8.313 × 10^{−5} | 2.472 × 10^{−5} |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Yang, J.; Cai, M.; Yang, X.; Zhou, Z.
Underwater Image Classification Algorithm Based on Convolutional Neural Network and Optimized Extreme Learning Machine. *J. Mar. Sci. Eng.* **2022**, *10*, 1841.
https://doi.org/10.3390/jmse10121841

**AMA Style**

Yang J, Cai M, Yang X, Zhou Z.
Underwater Image Classification Algorithm Based on Convolutional Neural Network and Optimized Extreme Learning Machine. *Journal of Marine Science and Engineering*. 2022; 10(12):1841.
https://doi.org/10.3390/jmse10121841

**Chicago/Turabian Style**

Yang, Junyi, Mudan Cai, Xingfan Yang, and Zhiyu Zhou.
2022. "Underwater Image Classification Algorithm Based on Convolutional Neural Network and Optimized Extreme Learning Machine" *Journal of Marine Science and Engineering* 10, no. 12: 1841.
https://doi.org/10.3390/jmse10121841