Font Size: a A A

The Information Fusion In Underwater Target Recognition

Posted on:2010-05-26Degree:MasterType:Thesis
Country:ChinaCandidate:Q GeFull Text:PDF
GTID:2178360272480371Subject:Naval Architecture and Marine Engineering
Abstract/Summary:PDF Full Text Request
The Intelligence Fusion in Underwater Target Recognition is based on the measurement of target vibration and noise, which aim is using the connatural characteristics to recognize the underwater target. After gained the comprehensive and reliable data, we extract the connatural characteristics of the underwater target by signal pretreatment, and then we have some new eigenvectors and finish the data fusion on characteristics level. This paper used the time domain, frequency domain and time-frequency domain signal progress method refines the character of underwater target, and then use the ANN classifier and show of hands method distinguish the two kinds of target. In the course of programming, it made use of the technique of blending C++ Builder and Matlab to improve code efficiency.The article can be divided into four parts: Introduction, feature extraction, information fusion and software programming.The introduction mainly discussed the background and significance of the paper, introduced the way of researching.The process of feature extraction is to transform the measure signal to different feature vectors which reflect the category of the input sample. The extracted features are the input modes to the classifier. Through the inspecting, the author selected three methods of the feature extraction in time-domain, frequency-domain and time-frequency- domain. Above all, It can fuse the different sides' feature and offer strong elements for the next process: classifying.The process of intelligence fusion is using fuzzy fusion classifier. It is using "feature extraction-classify" project to avoid the negative effect.The last part is software programming and data treating. To get the ratio of recognition and make a conclusion that the feature extraction arithmetic and classifier arithmetic are efficient, the article gave the conclusion by analyzed and deal with the data which of two kinds of experiment.
Keywords/Search Tags:feature extraction, feature fusion, artificial neural net work, decision level fusion
PDF Full Text Request
Related items