针对爆炸信号光电检测应用中出现噪声干扰影响目标识别的问题,分析研究了目标与干扰的形态差异,提出采用相关滤波法识别目标,并给出了该方法的理论最优解,结合目标特征构造了一种基于正弦波的M形模板;在此基础上构造含可变参数的模板,能够通过自学习获得目标信号的形态特征,并将学习结果用于该类目标的识别。对实测数据的处理表明,该方法可以有效解决噪声干扰严重时,系统无法正确识别目标的问题。

Aiming at the problem of noise interference affecting target recognition in the application of blast signal photoelectric detection, the morphological difference between target and interference is analyzed. The target filtering method is used to identify the target, and the theoretical optimal solution of the method is given. An M-shaped template based on sine wave is constructed. Based on this, a template with variable parameters is constructed. The morphological features of the target signal can be obtained through self-learning, and the learning result is used for the recognition of the target. The processing of the measured data shows that the method can effectively solve the problem that the system cannot correctly identify the target when the noise interference is serious.