A Multitarget Passive Recognition and Location Method Fusing SVM and BSS
A multitarget passive recognition and location method which fuses SVM and blind signal processing technique is proposed in this paper. Its characters are: Sampling data via multitarget information receiving array at first; And then getting separated signal and matrix by blind signal separation (BSS) to these data; Completing classification of each separated signal by using decision tree support vector machine (SVM) multitarget recognition process to the separated signal; Obtaining direction information of each signal by blind deconvolution location algorithm based on array model to the separated matrix at the same time; Finally, realizing target recognition and location by synthesizing targets information of the classification and direction. This paper studies technique principle of this method, gives a detailed implement step and proves its validity by multitarget recognition and location experiment of measured ship-radiated noise.
KeywordsSVM Blind signal Multitarget Recognition location
Unable to display preview. Download preview PDF.
- 2.Eriksson, J., Koivunen, V.: Complex-valued ICA using second order statistics. In: Proceedings of the 2004 IEEE Signal Processing Society Workshop on Machine Learning for Signal Processing 2004, pp. 183–191 (2004)Google Scholar
- 8.Gao, H., Liu, W.: An improved SVM classifier ICIC Express Letters 3(4), 1001–1005 (2009)Google Scholar