Comparison of Different Classification Algorithms for Underwater Target Discrimination

Li, Donghui; Azimi-Sadjadi, M.R.; Robinson, M.

doi:10.1109/tnn.2003.820621

Cited by 34 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The returning echo owns important frequency features information reflecting the target strength and the key point of underwater mine shell discrimination is robust echo features extraction process [10,14]. TF analyses, such as Wigner-Ville distribution, STFT, all produce poor results in impulse noise environment [7].…”

Section: Feature Extraction and Reductionmentioning

confidence: 99%

“…The canonical correlations are selected as features for classifying mine-like from non-minelike objects [17]. The authors of [14,19] demonstrated that the classification accuracy can be achieved largely when the backscattered signals' feature can be fused by some linear or nonlinear fusion schemes. Several classification algorithms of underwater targets from the acoustic backscattered signals are compared.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Underwater bottom still mine classification using robust time–frequency feature and relevance vector machine

Wang

2009

International Journal of Computer Mathematics

View full text Add to dashboard Cite

A novel approach to the problem of detecting and classifying underwater bottom mine objects in littoral environments from acoustic backscattered signals is considered. We begin by defining robust short-time Fourier transform to convert the received echo into a time-frequency (TF) plane. Identify interest local region in spectrogram, then features in TF plane with robustness to reverberation and noise disturbances are built. Finally, echo features are sent to a relevance vector machine (RVM) classifier that represents a Bayesian extension of support vector machine (SVM). To evaluate the performace of the classifier based on this approach, the classification experiment of two typical types of mines lying on the bottom have been performed with a broad bandwidth active sonar. Each of the targets was lying on the lake bottom at a depth of 20 m. The case study exploits the robustness of a feature extraction scheme, and furthermore, RVM yields a much sparser solution and improves the classification accuracy than SVM in an impulse noise environment.

show abstract

Section: Feature Extraction and Reductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Underwater bottom still mine classification using robust time–frequency feature and relevance vector machine

Wang

2009

International Journal of Computer Mathematics

View full text Add to dashboard Cite

show abstract

“…SVM provides a uniform framework to solve learning problem of few samples [5] . SVM classifier overcomes over-learning and local optimal value in NN method.…”

Section: Classifiersmentioning

confidence: 99%

Object Recognition Algorithm of Sonar Image

Tian

Zhou

Liu

2006

2006 8th International Conference on Signal Processing

View full text Add to dashboard Cite

In the target detection and recognition of underwater sonar image, object recognition is one of key technologies. With analysis and calculation, three features which have good distinguish degree are chosen to construct feature vector for classification, Clustering method, Neural network and Support vector machine were presented. Correct classification ratio can be above 95%. Simulation result indicates that features presented in this paper have characteristics such as precision and robustness, which reduce dependence of classifying result on classifier.

show abstract

“…The analysis of the acoustic backscattered signals of MMB and related substrates has been studied with a variety of different approaches [6][7][8][9]. One of them [9] has been the target for improvement in a work using GP [10], where GP was able to provide an improved discrimination between the different seafloor habitats.…”

Section: Introductionmentioning

confidence: 99%

Classification of Seafloor Habitats Using Genetic Programming

Silva

Tseng

2008

Lecture Notes in Computer Science

View full text Add to dashboard Cite

In this paper we use Genetic Programming for the classification of different seafloor habitats, based on the acoustic backscatter data from an echo sounder. By developing a different fitness function and dividing the multiple-class problem into several two-class problems, we were able to improve the results presented in a previously published work, providing a better discrimination between most of the seafloor types used in this study. We discuss the quality of these results and provide ideas to further improve the classification performance.

show abstract

Comparison of Different Classification Algorithms for Underwater Target Discrimination

Cited by 34 publications

References 12 publications

Underwater bottom still mine classification using robust time–frequency feature and relevance vector machine

Underwater bottom still mine classification using robust time–frequency feature and relevance vector machine

Object Recognition Algorithm of Sonar Image

Classification of Seafloor Habitats Using Genetic Programming

Contact Info

Product

Resources

About