On-Line Multi-Class Segmentation of Side-Scan Sonar Imagery Using an Autonomous Underwater Vehicle
<p>Side-scan sonar model. The <span class="html-italic">x</span> axis points to the AUV motion direction or along-track direction.</p> "> Figure 2
<p>An example of a swath composed of 500 bins. Bins from 0 to 249 come from the port sensing head. Bins from 250 to 499 are provided by the starboard sensing head.</p> "> Figure 3
<p>Example of acoustic image. Source: [<a href="#B37-jmse-08-00557" class="html-bibr">37</a>].</p> "> Figure 4
<p>Example of intensity and slant corrected acoustic image. The blind zone as well as the central bins, separating port and starboard, are outlined. The blind and low contrast zones under constant altitude and flat floor assumption are also shown.</p> "> Figure 5
<p>The Neural Network architecture.</p> "> Figure 6
<p>The training process.</p> "> Figure 7
<p>On-line usage of the NN and the Map Building (MB).</p> "> Figure 8
<p>Example of (<b>a</b>) a set of informative swaths and the corresponding segmented images using (<b>b</b>) SCM and (<b>c</b>) MCM.</p> "> Figure 9
<p>Trajectory followed by the AUV.</p> "> Figure 10
<p>Examples of the three considered classes: (<b>a</b>) rock, (<b>b</b>) rippled sand and (<b>c</b>,<b>d</b>) other.</p> "> Figure 11
<p>Example of data processing. (<b>a</b>,<b>b</b>): Modelled echo intensity <math display="inline"><semantics> <mrow> <mi>E</mi> <mo>(</mo> <mi>p</mi> <mo>)</mo> </mrow> </semantics></math> according to Equation (<a href="#FD5-jmse-08-00557" class="html-disp-formula">5</a>). (<b>c</b>,<b>d</b>): Raw SSS data. (<b>e</b>,<b>f</b>): Intensity and slant corrected acoustic image.</p> "> Figure 12
<p>F1-Scores for SCM training with (<b>a</b>) <math display="inline"><semantics> <mrow> <mi>p</mi> <mi>S</mi> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math>, (<b>b</b>) <math display="inline"><semantics> <mrow> <mi>p</mi> <mi>S</mi> <mo>=</mo> <mfrac> <mrow> <mi>N</mi> <mo>−</mo> <mn>1</mn> </mrow> <mn>2</mn> </mfrac> </mrow> </semantics></math> and (<b>c</b>) <math display="inline"><semantics> <mrow> <mi>p</mi> <mi>S</mi> <mo>=</mo> <mi>N</mi> </mrow> </semantics></math>.</p> "> Figure 13
<p>F1-Scores for MCM training with (<b>a</b>) <math display="inline"><semantics> <mrow> <mi>p</mi> <mi>S</mi> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math>, (<b>b</b>) <math display="inline"><semantics> <mrow> <mi>p</mi> <mi>S</mi> <mo>=</mo> <mfrac> <mrow> <mi>N</mi> <mo>−</mo> <mn>1</mn> </mrow> <mn>2</mn> </mfrac> </mrow> </semantics></math> and (<b>c</b>) <math display="inline"><semantics> <mrow> <mi>p</mi> <mi>S</mi> <mo>=</mo> <mi>N</mi> </mrow> </semantics></math>.</p> "> Figure 14
<p>Execution times for (<b>a</b>) training, (<b>b</b>) segmenting using SCM and (<b>c</b>) segmenting using MCM.</p> "> Figure 15
<p>Segmentation results. (<b>a</b>) Informative images corresponding to a small transect overlaid with the ground truth and segmented images using (<b>b</b>) SCM and <math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>t</mi> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math>, (<b>c</b>) MCM and <math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>t</mi> <mo>=</mo> <mn>1</mn> </mrow> </semantics></math>, (<b>d</b>) SCM and <math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>t</mi> <mo>=</mo> <mn>41</mn> </mrow> </semantics></math>, (<b>e</b>) MCM and <math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>t</mi> <mo>=</mo> <mn>41</mn> </mrow> </semantics></math>, (<b>f</b>) SCM and <math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>t</mi> <mo>=</mo> <mn>83</mn> </mrow> </semantics></math> and (<b>g</b>) MCM and <math display="inline"><semantics> <mrow> <mo>Δ</mo> <mi>t</mi> <mo>=</mo> <mn>83</mn> </mrow> </semantics></math>.</p> ">
Abstract
:1. Introduction
- Derive an acoustics based method [17] to pre-process the data so that the NN has to deal with less uncertainties, thus facilitating its training and on-line usage.
- Propose a sliding window approach that makes it possible, when combined with the pre-processing, to train the NN with a small amount of data and to use it on-line even on AUVs with reduced computational power.
- Propose a Convolutional Neural Network following an encoder-decoder architecture in charge of segmenting the acoustic data.
2. The Side-Scan Sonar
2.1. Overview
2.2. Sensor Operation
2.3. Acoustic Image Formation
3. Data Pre-Processing
3.1. Overview
3.2. Intensity Correction
3.3. Slant Range Correction
3.4. Data Selection
4. Data Segmentation
4.1. Overview
4.2. The Neural Network
4.2.1. Training
4.2.2. On-Line Usage
4.3. Map Building
5. Experimental Results
5.1. Overview
5.2. System Parametrization
5.3. Quantitative Results
5.4. Qualitative Results
5.5. Discussion
6. Conclusions and Future Work
Author Contributions
Funding
Acknowledgments
Conflicts of Interest
References
- Burguera, A.B.; Bonin-Font, F. A trajectory-based approach to multi-session underwater visual SLAM using global image signatures. J. Mar. Sci. Eng. 2019, 7, 278. [Google Scholar] [CrossRef] [Green Version]
- Köser, K.; Frese, U. Challenges in Underwater Visual Navigation and SLAM. In Intelligent Systems, Control and Automation: Science and Engineering; Springer: Cham, Switzerland, 2020; Volume 96, pp. 125–135. [Google Scholar] [CrossRef]
- Wu, Y.; Ta, X.; Xiao, R.; Wei, Y.; An, D.; Li, D. Survey of underwater robot positioning navigation. Appl. Ocean Res. 2019. [Google Scholar] [CrossRef]
- Marage, J.P.; Mori, Y. Sonar and Underwater Acoustics; John Wiley & Sons Ltd.: Hoboken, NJ, USA, 2013. [Google Scholar] [CrossRef]
- Le Bas, T.P.; Somers, M.L.; Campbell, J.M.; Beale, R. Swath bathymetry with GLORIA. IEEE J. Ocean. Eng. 1996, 21, 545–552. [Google Scholar] [CrossRef]
- Searle, R.C.; Le Bas, T.P.; Mitchell, N.C.; Somers, M.L.; Parson, L.M.; Patriat, P. GLORIA image processing: The state of the art. Mar. Geophys. Res. 1990, 12, 21–39. [Google Scholar] [CrossRef]
- Burguera, A. Underwater Localization using Probabilistic Sonar Registration and Pose Graph Optimization. In Proceedings of the 2018 IEEE/OES Autonomous Underwater Vehicle Workshop (AUV), Porto, Portugal, 6–9 November 2018. [Google Scholar] [CrossRef]
- Marx, D.; Nelson, M.; Chang, E.; Gillespie, W.; Putney, A.; Warman, K. Introduction to synthetic aperture sonar. IEEE Signal Process. Workshop Stat. Signal Array Process. SSAP 2000, 717–721. [Google Scholar] [CrossRef] [Green Version]
- Ribas, D.; Ridao, P.; Neira, J. Understanding Mechanically Scanned Imaging Sonars. In Underwater SLAM for Structured Environments Using an Imaging Sonar; Springer: Berlin, Germany, 2010; pp. 37–46. [Google Scholar] [CrossRef]
- Sousa-Sena, A.L. Shallow Water Remote Sensing Using Sonar Improved with Geostatistics and Stochastic Resonance Data Processing. Ph.D. Thesis, Universitat de les Illes Balears, Palma, Illes Balears, Spain, 2018. [Google Scholar]
- Ji, D.; Liu, J. Multi-Beam Sonar Application on Autonomous Underwater Robot. Mar. Geod. 2015, 38, 281–288. [Google Scholar] [CrossRef]
- Van Veen, B.D.; Buckley, K.M. Beamforming: A Versatile Approach to Spatial Filtering. IEEE ASSP Mag. 1988, 5, 4–24. [Google Scholar] [CrossRef]
- Mallios, A.; Vidal, E.; Campos, R.; Carreras, M. Underwater caves sonar data set. Int. J. Robot. Res. 2017, 36, 1247–1251. [Google Scholar] [CrossRef]
- Jiang, M.; Song, S.; Li, Y.; Jin, W.; Liu, J.; Feng, X. A Survey of Underwater Acoustic SLAM System. In Lecture Notes in Computer Science; Springer: Cham, Switzerland, 2019; Volume 11741 LNAI, pp. 159–170. [Google Scholar] [CrossRef]
- Sternlicht, D.D. Historical development of side scan sonar. J. Acoust. Soc. Am. 2017, 141, 4041. [Google Scholar] [CrossRef]
- Savini, A. Side-Scan Sonar as a Tool for Seafloor Imagery: Examples from the Mediterranean Continental Margin. In Sonar Systems; IntechOpen Ltd.: London, UK, 2011. [Google Scholar] [CrossRef] [Green Version]
- Burguera, A.; Oliver, G. High-resolution underwater mapping using Side-Scan Sonar. PLoS ONE 2016, 11. [Google Scholar] [CrossRef] [Green Version]
- Johnson, H.P.; Helferty, M. The geological interpretation of Side-Scan Sonar. Rev. Geophys. 1990. [Google Scholar] [CrossRef] [Green Version]
- Bava-De-Camargo, P.F. The use of side scan sonar in Brazilian Underwater Archaeology. In Proceedings of the IEEE/OES Acoustics in Underwater Geosciences Symposium, Rio de Janeiro, Brazil, 29–31 July 2016. [Google Scholar] [CrossRef]
- Cobra, D.T.; Oppenheim, A.V.; Jaffe, J.S. Geometric Distortions in Side-Scan Sonar Images: A Procedure for Their Estimation and Correction. IEEE J. Ocean. Eng. 1992, 17, 252–268. [Google Scholar] [CrossRef]
- Sheffer, T.; Guterman, H. Geometrical Correction of Side-scan Sonar Images. In Proceedings of the 2018 IEEE International Conference on the Science of Electrical Engineering in Israel, ICSEE 2018, Eilat, Israel, 12–14 December 2018. [Google Scholar] [CrossRef]
- Bikonis, K.; Moszynski, M.; Lubniewski, Z. Application of shape from shading technique for side scan sonar images. Pol. Marit. Res. 2013, 20, 39–44. [Google Scholar] [CrossRef] [Green Version]
- Reed, S.; Petillot, Y.; Bell, J. Mine detection and classification in side scan sonar. Sea Technol. 2004, 45, 35–39. [Google Scholar]
- Aulinas, J.; Lladó, X.; Salvi, J.; Petillot, Y.R. Feature based SLAM using side-scan salient objects. In Proceedings of the MTS/IEEE OCEANS, Seattle, WA, USA, 20–23 September 2010. [Google Scholar] [CrossRef]
- Moreno, D.; Burguera, A.; Oliver, G. SSS-SLAM: An Object Oriented Matlab Framework for Underwater SLAM using Side Scan Sonar. In Proceedings of the XXXV Jornadas de Automática, Valencia, Spain, 3–5 September 2014. [Google Scholar]
- Saini, K.; Dewal, M.L.; Rohit, M. Ultrasound Imaging and Image Segmentation in the area of Ultrasound: A Review. Int. J. Adv. Sci. Technol. 2010, 24, 41–60. [Google Scholar]
- Priyadharsini, R.; Sharmila, T.S. Object Detection in Underwater Acoustic Images Using Edge Based Segmentation Method. Procedia Comput. Sci. 2019, 165, 759–765. [Google Scholar] [CrossRef]
- Williams, D.P. Fast Unsupervised Seafloor Characterization in Sonar Imagery Using Lacunarity. IEEE Trans. Geosci. Remote Sens. 2015, 53, 6022–6034. [Google Scholar] [CrossRef]
- Khidkikar, M.; Balasubramanian, R. Segmentation and classification of side-scan sonar data. In Proceedings of the Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Montreal, QC, Canada, 3–5 October 2012; Volume 7506 LNAI, pp. 367–376. [Google Scholar] [CrossRef]
- Pinto, M.; Ferreira, B.; Matos, A.; Cruz, N. Side scan sonar image segmentation and feature extraction. In Proceedings of the MTS/IEEE Biloxi—Marine Technology for Our Future: Global and Local Challenges, OCEANS 2009, Biloxi, MS, USA, 26–29 October 2009. [Google Scholar]
- Daniel, S.; Le Léannec, F.; Roux, C.; Solaiman, B.; Maillard, E.P. Side-scan sonar image matching. IEEE J. Ocean. Eng. 1998, 23, 245–259. [Google Scholar] [CrossRef]
- Sultana, F.; Sufian, A.; Dutta, P. Evolution of Image Segmentation using Deep Convolutional Neural Network: A Survey. Knowl.-Based Syst. 2020, 201–202. [Google Scholar] [CrossRef]
- Alhasoun, F.; Gonzalez, M. Streetify: Using Street View Imagery and Deep Learning for Urban Streets Development. In Proceedings of the IEEE International Conference on Big Data, Los Angeles, CA, USA, 9–12 December 2019; pp. 2001–2006. [Google Scholar] [CrossRef] [Green Version]
- Van Opbroek, A.; Achterberg, H.C.; Vernooij, M.W.; De Bruijne, M. Transfer learning for image segmentation by combining image weighting and kernel learning. IEEE Trans. Med. Imaging 2019, 38, 213–224. [Google Scholar] [CrossRef] [Green Version]
- Yu, F.; Zhu, Y.; Wang, Q.; Li, K.; Wu, M.; Li, G.; Yan, T.; He, B. Segmentation of Side Scan Sonar Images on AUV. In Proceedings of the 2019 IEEE Underwater Technology (UT), Kaohsiung, Taiwan, 16–19 April 2019; pp. 1–4. [Google Scholar] [CrossRef]
- Coiras, E.; Petillot, Y.; Lane, D.M. Multiresolution 3-D reconstruction from side-scan sonar images. IEEE Trans. Image Process. 2007, 16, 382–390. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Burguera, A. Segmentation of Side-Scan Sonar Data—Source Code. Available online: https://github.com/aburguera/NNSSS (accessed on 16 July 2020).
- Burguera, A. A novel approach to register sonar data for underwater robot localization. In Proceedings of the Intelligent Systems Conference (IntelliSys 2017), London, UK, 7–8 September 2017; Volume 2018-January, pp. 1034–1043. [Google Scholar] [CrossRef]
- Chang, Y.C.; Hsu, S.K.; Tsai, C.H. Sidescan sonar image processing: Correcting brightness variation and patching gaps. J. Mar. Sci. Technol. 2010, 18, 785–789. [Google Scholar]
- Kleeman, L.; Kuc, R. Sonar Sensing. In Springer Handbook of Robotics; Springer: Berlin/Heidelberg, Germany, 2008; pp. 491–519. [Google Scholar] [CrossRef]
- Greenspan, M.; Tschiegg, C.E. Tables of the Speed of Sound in Water. J. Acoust. Soc. Am. 1959, 31, 75–76. [Google Scholar] [CrossRef]
- Abramowitz, M.; Stegun, I.A. Handbook of Mathematical Functions: With Formulas, Graphs, and Mathematical Tables; Applied Mathematics Series 55; U.S. Government Printing Office: Washington, DC, USA, 1964; pp. 591–592.
- Langer, D.; Hebert, M. Building qualitative elevation maps from side scan sonar data for autonomous underwater navigation. In Proceedings of the 1991 IEEE International Conference on Robotics and Automation, Sacramento, CA, USA, 9–11 April 1991; Volume 3, pp. 2478–2483. [Google Scholar] [CrossRef] [Green Version]
- Chollet, F. Deep Learning with Phyton; Manning Publications Co.: Shelter Island, NY, USA, 2018; p. 386. [Google Scholar]
30 | |
---|---|
3 | |
20 | |
f | 800 KHz |
1.95 mm | |
30 m | |
0.12 m | |
Bins per swath | 250 port, 250 starboard |
Sampling frequency | 10 swath/s |
Transect | Swaths | Rock | Sand | Other |
---|---|---|---|---|
1 | 5764 | 125,294 (13.095%) | 26,464 (2.766%) | 805,066 (84.139%) |
2 | 6800 | 107,706 (9.542%) | 133,665 (11.841%) | 887,429 (78.617%) |
3 | 3825 | 253,029 (39.850%) | 111,411 (17.546%) | 270,510 (42.603%) |
4 | 3517 | 215,148 (36.852%) | 70,604 (12.093%) | 298,070 (51.055%) |
5 | 2532 | 136,762 (32.538%) | 38,650 (9.196%) | 244,900 (58.266%) |
GLOBAL | 22,438 | 837,939 (22.497%) | 380,794 (10.223%) | 2,505,975 (67.280%) |
(a) | ||||
---|---|---|---|---|
True | Rock | Sand | Other | |
Pred. | ||||
Rock | 0.7917 | 0.0347 | 0.0428 | |
Sand | 0.1625 | 0.9409 | 0.1308 | |
Other | 0.0457 | 0.0242 | 0.8263 | |
(b) | ||||
True | Rock | Sand | Other | |
Pred. | ||||
Rock | 0.8803 | 0.1022 | 0.0174 | |
Sand | 0.0602 | 0.9221 | 0.0177 | |
Other | 0.1111 | 0.1558 | 0.7331 |
(a) | ||||
---|---|---|---|---|
True | Rock | Sand | Other | |
Pred. | ||||
Rock | 0.8015 | 0.0367 | 0.0461 | |
Sand | 0.1554 | 0.9399 | 0.1358 | |
Other | 0.0430 | 0.0233 | 0.8180 | |
(b) | ||||
True | Rock | Sand | Other | |
Pred. | ||||
Rock | 0.8721 | 0.1085 | 0.0193 | |
Sand | 0.0563 | 0.9247 | 0.0189 | |
Other | 0.1021 | 0.1502 | 0.7476 |
ACCURACY (SCM) | ||||
---|---|---|---|---|
1 | ||||
pS | ||||
1 | 0.9100 | 0.9020 | 0.9020 | |
0.8990 | 0.8950 | 0.8930 | ||
0.8830 | 0.8770 | 0.8790 | ||
ACCURACY (MCM) | ||||
1 | ||||
pS | ||||
1 | 0.9100 | 0.9070 | 0.9020 | |
0.8990 | 0.8970 | 0.8930 | ||
0.8840 | 0.8820 | 0.8790 |
Precision | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.8860 | 0.9350 | 0.7960 | 0.9050 | 0.9250 | 0.7430 | 0.8720 | 0.9320 | 0.7780 | |
0.8560 | 0.9370 | 0.7430 | 0.8770 | 0.9310 | 0.6960 | 0.8430 | 0.9360 | 0.7240 | ||
0.8950 | 0.9030 | 0.7290 | 0.9080 | 0.8970 | 0.6740 | 0.8810 | 0.9030 | 0.7150 | ||
Recall | ||||||||||
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.8300 | 0.9420 | 0.8850 | 0.7940 | 0.9430 | 0.9030 | 0.8190 | 0.9360 | 0.8710 | |
0.8270 | 0.9340 | 0.8190 | 0.8000 | 0.9360 | 0.8450 | 0.8170 | 0.9290 | 0.8190 | ||
0.7630 | 0.9510 | 0.7640 | 0.7340 | 0.9510 | 0.7910 | 0.7570 | 0.9470 | 0.7530 | ||
Fall-Out | ||||||||||
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.0340 | 0.1310 | 0.0230 | 0.0280 | 0.1480 | 0.0290 | 0.0380 | 0.1390 | 0.0250 | |
0.0420 | 0.1310 | 0.0290 | 0.0370 | 0.1410 | 0.0340 | 0.0460 | 0.1350 | 0.0310 | ||
0.0320 | 0.1810 | 0.0310 | 0.0280 | 0.1890 | 0.0370 | 0.0360 | 0.1830 | 0.0330 | ||
F1-Score | ||||||||||
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.8570 | 0.9390 | 0.8380 | 0.8460 | 0.9340 | 0.8160 | 0.8450 | 0.9340 | 0.8220 | |
0.8410 | 0.9350 | 0.7790 | 0.8360 | 0.9330 | 0.7630 | 0.8300 | 0.9320 | 0.7690 | ||
0.8240 | 0.9270 | 0.7460 | 0.8120 | 0.9230 | 0.7280 | 0.8140 | 0.9240 | 0.7340 |
Precision | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.8870 | 0.9360 | 0.7950 | 0.8780 | 0.9340 | 0.7890 | 0.8720 | 0.9320 | 0.7780 | |
0.8550 | 0.9380 | 0.7400 | 0.8520 | 0.9370 | 0.7370 | 0.8430 | 0.9360 | 0.7240 | ||
0.8930 | 0.9040 | 0.7280 | 0.8880 | 0.9040 | 0.7230 | 0.8810 | 0.9030 | 0.7150 | ||
Recall | ||||||||||
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.8300 | 0.9420 | 0.8870 | 0.8240 | 0.9400 | 0.8780 | 0.8190 | 0.9360 | 0.8710 | |
0.8280 | 0.9340 | 0.8210 | 0.8260 | 0.9320 | 0.8190 | 0.8170 | 0.9290 | 0.8190 | ||
0.7650 | 0.9510 | 0.7650 | 0.7620 | 0.9490 | 0.7590 | 0.7570 | 0.9470 | 0.7530 | ||
Fall-Out | ||||||||||
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.0330 | 0.1310 | 0.0230 | 0.0360 | 0.1340 | 0.0240 | 0.0380 | 0.1390 | 0.0250 | |
0.0420 | 0.1300 | 0.0290 | 0.0430 | 0.1320 | 0.0300 | 0.0460 | 0.1350 | 0.0310 | ||
0.0330 | 0.1790 | 0.0310 | 0.0340 | 0.1800 | 0.0320 | 0.0360 | 0.1830 | 0.0330 | ||
F1-Score | ||||||||||
1 | ||||||||||
PS | Rock | Sand | Other | Rock | Sand | Other | Rock | Sand | Other | |
1 | 0.8570 | 0.9390 | 0.8390 | 0.8500 | 0.9370 | 0.8310 | 0.8450 | 0.9340 | 0.8220 | |
0.8410 | 0.9360 | 0.7780 | 0.8380 | 0.9350 | 0.7760 | 0.8300 | 0.9320 | 0.7690 | ||
0.8240 | 0.9270 | 0.7460 | 0.8200 | 0.9260 | 0.7400 | 0.8140 | 0.9240 | 0.7340 |
Time consumption (SCM) | |||||||
---|---|---|---|---|---|---|---|
1 | |||||||
PS | Training | Segment. | Training | Segment. | Training | Segment. | |
1 | 1373.2110 ms | 3.2084 ms | 1373.2110 ms | 0.2174 ms | 1373.2110 ms | 0.2148 ms | |
36.0540 ms | 3.2304 ms | 36.0540 ms | 0.2295 ms | 36.0540 ms | 0.2294 ms | ||
18.0790 ms | 3.3044 ms | 18.0790 ms | 0.2323 ms | 18.0790 ms | 0.2145 ms | ||
Time consumption (MCM) | |||||||
1 | |||||||
PS | Training | Segment. | Training | Segment. | Training | Segment. | |
1 | 1373.2110 ms | 3.8411 ms | 1373.2110 ms | 0.6879 ms | 1373.2110 ms | 0.6580 ms | |
36.0540 ms | 3.8038 ms | 36.0540 ms | 0.6799 ms | 36.0540 ms | 0.6163 ms | ||
18.0790 ms | 3.7451 ms | 18.0790 ms | 0.6960 ms | 18.0790 ms | 0.6328 ms |
© 2020 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 (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Burguera, A.; Bonin-Font, F. On-Line Multi-Class Segmentation of Side-Scan Sonar Imagery Using an Autonomous Underwater Vehicle. J. Mar. Sci. Eng. 2020, 8, 557. https://doi.org/10.3390/jmse8080557
Burguera A, Bonin-Font F. On-Line Multi-Class Segmentation of Side-Scan Sonar Imagery Using an Autonomous Underwater Vehicle. Journal of Marine Science and Engineering. 2020; 8(8):557. https://doi.org/10.3390/jmse8080557
Chicago/Turabian StyleBurguera, Antoni, and Francisco Bonin-Font. 2020. "On-Line Multi-Class Segmentation of Side-Scan Sonar Imagery Using an Autonomous Underwater Vehicle" Journal of Marine Science and Engineering 8, no. 8: 557. https://doi.org/10.3390/jmse8080557