一种基于线谱特征函数提取LOFAR图线谱的方法

发布时间：2018-02-03 16:19

本文关键词： LOFAR图线谱特征被动检测线谱特征累积检测法　出处：《声学技术》2016年04期 　论文类型：期刊论文

【摘要】：水声信号被动检测中广泛使用LOFAR图对接收信号进行处理和分析。针对LOFAR图中线谱信号检测问题,根据线谱信号特征设计特征函数,提出频域滑动窗线谱特征累积检测法。该方法在频率轴移动观察窗,用多步决策算法计算每个观察窗的最优解,得到最优路径,如果最优路径特征值大于阈值,则累积LOFAR图像素点被该最优路径经过的次数,次数越多对应点为线谱点的概率越大。仿真研究表明,该方法对频率时变、低信噪比的线谱信号具有良好的检测能力,可实现多根线谱的增强与检测。海试数据处理结果证明了该方法的可行性和稳健性。该算法对于辐射线谱信号的水下目标远距离探测识别有较高的参考价值。
[Abstract]:In the passive detection of underwater acoustic signal, LOFAR diagram is widely used to process and analyze the received signal. To solve the problem of line spectrum signal detection in LOFAR diagram, the characteristic function is designed according to the characteristics of line spectrum signal. In this paper, a method of frequency domain sliding window line spectrum feature accumulation detection is proposed, which moves the observation window in the frequency axis, and calculates the optimal solution of each window by using a multi-step decision algorithm, and obtains the optimal path, if the eigenvalue of the optimal path is greater than the threshold. Then the number of pixel points passing through the optimal path and the probability that the number of pixels passing through the optimal path is greater. Simulation results show that the method has a good ability to detect linear spectrum signals with time-varying frequency and low signal-to-noise ratio (SNR). The results of sea trial data processing prove the feasibility and robustness of the method. The algorithm has a high reference value for underwater target detection and identification of radiation line spectrum signals.
【作者单位】：中国科学院声学研究所声场声信息国家重点实验室;中国科学院大学;
【基金】：国家自然科学基金资助项目(11434012,41561144006)
【分类号】：TB56
【正文快照】： 0引言水声被动探测和目标识别领域中,线谱信号由于具有相对较高的信噪比并携带声源特征信息,长期以来一直被广泛研究和应用于水声工程实践。如吴国清等利用谱峰形状特点,给出了一套线谱识别逻辑[1];陈敬军等提出模仿声呐员识别线谱过程的基于人工智能的线谱检测技术[2];陈阳等

【相似文献】