Analysis And Extracting Of The Characteristic Of Acoustic Backscatter

Detection by active sonar is one of the most important measures for underwater exploration. But shallow sea environment is indeed complex, especially like reverberation, so it is a very difficult task to detect and classify a target. Research are carried in the direction of developping an approach for the detection and classification of submerged elastic objects.The theory of resonance scattering of elastic object is introduced firstly. Then the possibility analysis of identifying an elastic target based on RST is discussed. A plan which includes time delay estimation algorithm and adaptive filter algorithm is devised to extract the resonance scattering under the present experiment condition. Test results indicate the confirmation of elastic resonance. Besides these, another means including deconvolution and AR model spectrum estimation is attempted to estimate the resonance scattering.At the second part of the paper, Time-Frequency analysis applied below are discussed. Several algorithms of weakening and removing cross term are investigated and proved effective by emulator. At last, result and discussion of Song-Hua lake data processing are presented which show that Time-Frequency analysis can be a propose way to detect signal echo under a certain Signal-to-Noise.