关键词: 倒谱分析 RASTA滤波 信号重构 水下目标识别

Abstract:

The underwater target＇s radiated noise posed through the convolution of the excitation source transferred and diffused by the target＇s body and the  signal which is captured by passive signal produced periodically by the sonar can be considered to be corntarget itself, and the radiated signal transfo  rm function of the oceanic acoustic channel, To recognize the underwater targets, the first step is to filter the hash. This paper achieves this goal th  rough homomorphic analysis. First of all, the relationships between the components of noise are converted to linear summarization through a cepstrum tra  nsform. Then the components of noise and the oceanic acoustic channel are separated through the time window and filtering. These processes are helpful i n distilling the signal features and to classify the targets. Experiments show that the proposed signal reconstruction method can precisely recover the   original signal. This will be convenient for the forthcoming feature extraction step and the classification step.

Key words: （cepstrum analysis, RASTA filtering, signal reconstruction, underwater target recognition）