Fast recognition algorithm of underwater micro-terrain based on ultrasonic detection

Abstract

An algorithm was proposed to fast recognize three types of underwater micro-terrain, i.e. the level, the gradient and the uneven. With pendulum single beam bathymeter, the hard level concrete floor, the random uneven floor and the gradient wood panel (8°) were ultrasonically detected 20 times, respectively. The results show that the algorithm is right from fact that the first clustering values of the uneven are all less than the threshold value of 60.0% that is obtained by the level and gradient samples. The algorithm based on the dynamic clustering theory can effectively eliminate the influences of the exceptional elevation values produced by the disturbances resulted from the grazing angle, the characteristic of bottom material and environmental noises, and its real-time capability is good. Thus, the algorithm provides a foundation for the next restructuring of the micro-terrain.

Keywords

underwater micro-terrain / ultrasonic detection / single beam / exceptional elevation values / threshold value / dynamic clustering

References

[1] XiaY.-m., BuY.-y., MaZ.-g., ZhaoH.-m., LuoB.-wen.. Modeling and simulation of ocean mining subsystem based on virtual prototyping technology [J]. Journal of Central South University of Technology, 2005, 12(2): 176-179

[2] SteinvallO. K., KoppariK. R., KarlssonU. C.. Experimental evaluation of an airborne depth-sounding lidar [J]. Optical Engineering, 1993, 32(6): 1307-1321

[3] QinX.-y., BuY.-yong.. Study on multilevel model of optimized cut depth in real-time mining of ocean cobalt-rich crusts [J]. Ocean Engineering, 2005, 23(3): 99-104

[4] YamazakiT., SharmaR.. Morphological features of Co-rich manganese deposits and their relation to seabed slopes [J]. Marine Georesources and Geotechnology, 2000, 18(1): 43-76

[5] RenZ.-l., HuangY.-ying.. A new method for underwater electromagnetic passive ranging [J]. J Huazhong Univ Sci Tech, 2001, 29(5): 89-91

[6] LiuJ.Study of micro-terrain detection system of single-beam based on fractal theory [D], 2006, Changsha, Central South University of Technology

[7] AntoniolG., CeccarelliM., MarateaA., RussoF.. Classification of digital terrain models through fuzzy clustering: An application [C]. Proceedings of 5th International Workshop on Fuzzy Logic and Application, 2005, Changsha, Springer-Verlag: 174-182

[8] HizirS., MuzailinA., KhaledB.. The application of fuzzy clustering to satellite images data [J]. Statistical Methods for Biostatistics and Related Fields, 2006, 11: 357-366

[9] MaJ., ShaoL.. An optimal algorithm for fuzzy classification problem [J]. Journal of Software, 2001, 12(4): 578-581

[10] RoubosH., SetnesM.. Compact and transparent fuzzy models and classifiers through iterative complexity reduction [J]. IEEE Trans Fuzzy Systems, 2001, 9(4): 516-524

[11] AbeS., LanM. S.. A method for fuzzy rules extraction directly from numerical data and its application to pattern classification [J]. IEEE Trans Fuzzy Systems, 1995, 3(1): 18-28

[12] PalN. R., BezdkJ. C.. On cluster validity for the fuzzy c-mean model [J]. IEEE Trans Fuzzy Systems, 1995, 3(3): 370-379

[13] MinS. H., IngooH.. Dynamic fuzzy clustering for recommender systems [C]. Proceedings of 9th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Ming, 2005, Changsha, Springer-Verlag: 480-485

[14] HoS. Y., LiuC. C., LiuS.. Design of an optimal nearest neighbor classifier using an intelligent genetic algorithm [J]. Pattern Recognition Letters, 2002, 23(13): 1495-1503

[15] Abdel-DayemA. R., Ei-SakkaM. R.. Fuzzy c-means clustering for segmenting carotid artery ultrasound images [C]. Proceedings of 4th International Conference on Image Analysis and Recognition, 2007, Changsha, Springer-Verlag: 945-948

[16] ZhaiY.-m., ZhaoR.-xing.. K-nearest neighbor nonparametric estimation bootstrap model for weather probability forecasting [J]. Journal of System Simulation, 2005, 17(4): 786-788